Special Publication 13, Frontiers of Karst Research

Special Publication 13

��

Frontiers of Karst Research Proceedings and recommendations of the workshop held May 3 through 5, 2007, San Antonio, Texas, USA 2008

Edited by Jonathan B. Martin William B. White Copyright © 2008 by the Karst Waters Institute, Inc. except where individual contributors to this volume retain copyright. All rights reserved with the exception of non-commercial photocopying for the purposes of scientific or educational advancement.

Published by: Karst Waters Institute, Inc. P.O. Box 4142 Leesburg, Virginia 20177 http://www.karstwaters.org

Please visit our web page for ordering information.

The Karst Waters Institute is a non-profit 501 (c) (3) research and education organization incorporated in West Virginia. The mission of the Institute is improvement of fundamental understanding of karst water systems through sound scientific research and the education of professionals and the public.

Library of Congress Control Number: 2008938009

ISBN number 978-0-9789976-2-5

Recommended citation for this volume:

Martin, J.B. and White, W.B. (eds.), 2008, Frontiers of karst research. Special Publication 13, Karst Waters Institute, Leesburg, Virginia

Printed in the U.S.A. by BookMasters, Inc., Ashland, Ohio Text: Times 9.5 pt on 70# white matte.

Cover Photo: Cylindrical stalagmites are rapidly becoming an important paleoclimate archive. This display of these speleothems is from Virgin Cave, Guadalupe Mountains, New Mexico. Photo by David Bunnell/Under Earth Images. WORKSHOP SUPPORTERS

The Karst Waters Institute greatly appreciates the financial and logistical support of the following agencies and organizations

National Science Foundation U.S. Army Research Office National Cave and Karst Research Institute University of Florida, Department of Geological Sciences

WORKSHOP ORGANIZERS AND STAFF

Workshop Co-Chairs Jonathan B. Martin Diana E. Northup Annette Summers-Engel William B. White

Book Production Editors Elaine L. Butcher Elizabeth L. White

FOCUS GROUP LEADERS

Jay L. Banner, Ph.D., University of Texas at Austin David C. Culver, Ph.D., American University Diana E. Northup, Ph.D., University of New Mexico Megan Porter, Ph.D., University of Maryland Baltimore County Martin Sauter, Ph.D., Universität Göttingen Geary Schindel, M.S., Edward Aquifer District Kevin Simon, Ph.D., University of Maine Annette Summers-Engel, Ph.D., Louisiana State University Dorothy J. Vesper, Ph.D., West Virginia University

CROSS-DISCIPLINARY REPORTERS

Penelope J. Boston, Ph.D., New Mexico Institute of Technology William B. White, Ph.D., The Pennsylvania State University KARST WATERS INSTITUTE OFFICERS FOR 2007

President Daniel W. Fong, Ph.D. Executive Vice-President William B. White, Ph.D. Treasurer David C. Culver, Ph.D. Secretary Diana E. Northup, Ph.D Vice President for Communications Ira D. Sasowsky, Ph.D. Vice President for Education Horton H. Hobbs, III, Ph.D. Vice President for Research Carol M. Wicks, Ph.D.

KARST WATERS INSTITUTE BOARD OF DIRECTORS, 2007

William K. Jones (Chair) John E. Mylroie, Ph.D. Robert N. Cronk Diana E. Northup, Ph.D. Emily Davis Ira D. Sasowsky, Ph.D. Janine Gibert, Ph.D. Annette Summers-Engel, Ph.D. Christopher G. Groves, Ph.D. H. Len Vacher, Ph.D. Horton H. Hobbs, III, Ph.D. William B. White, Ph.D. Philip M. LaMoreaux, Ph.D. Carol M. Wicks, Ph.D. Jonathan B.Martin, Ph.D. TABLE OF CONTENTS

Preface ...... vii

Part 1 – Frontiers of Karst Research ...... 1

Frontiers of Karst Research Opportunities and Recommendations: The Next Frontier ...... 3

Part 2 – Todays Frontier ...... 11

Modeling Karst Hydrodynamics. Kovács and Sauter ...... 13 Geochemistry and Climate Changes Banner, Musgrove, Rasmussen, Partin, Long, Katz, Mahler, Edwards, Cobb, James, Harmon, Herman, and Wicks ...... 27 Caves and Karst as Model Systems for Advancing the Microbial Sciences Engel and Northup ...... 37 Ecosystem Science and Karst Systems Simon...... 49 The Struggle to Measure Subterranean Biodiversity Culver...... 54 Evolution in Karst: Lineages, Ages, and Adaptation of Cave Faunas Porter...... 60 Karst Resources and Other Applied Issues Vesper ...... 65

Part 3 – Findings and Recommendations of the Focus Groups...... 75 Focus Group on Karst Hydrology - Conceptual Models, Aquifer Characterization, and Numerical Modeling Sauter, Covington, Florea, Gabrovsek, Gao, Green, Gulley, Harmon, Herman, Jeannin, Jones, Kincaid, Moore, Mylroie, Sasowsky, Screaton and Wicks...... 77 Focus Group on Geochemistry and Climate Banner, Boston, Colucci, Cowan, Frappier, Gentry, Harmon, Katz, Long, Martin, Musgrove, Partin, Rasmussen, Wong and White...... 82 Focus Group on Caves and Karst as Model Systems in Geomicrobiology Engel, Northup, Gary, Gonzalez, B., Gonzales, J., Hutchens, Jones, Macalady, Spear and Spilde...... 90 Focus Group on Ecosystem Function Simon, Fong, Hinderstein, Maloney, Payn, Vernarsky and Wilhelm...... 96 Focus Group on Subterranean Biodiversity Culver, Boston, Christman, Collins, Godwin, Hobbs, Holmes, Holsinger, Iliffe, Krejca, Lewis, O’Conner, Pipan, Schneider, Taylor and Zagmajster...... 98 Understanding the Tempo and Mode of Evolution: Cave Adaptation as a Model System Porter, Dittmar, Hutchins, Jeffery, Lefébure, Paquin and Protas...... 100 Focus Group on Karst Resources and Other Applied Issues Vesper, Schindel, Beck, Van Brahana, Cate, Engler, Ewers, Falkenberg, Halihan, Idstein, Krothe, Peterson, Toran, Veni, White and Williams...... 106

References for Focus Groups...... 108

Appendix: Workshop Participants...... 113

Frontiers of Karst Research v

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13

PREFACE

Some of the world's most exotic landscapes and rare and en- were broadly divided into topics designated as Hydrologic dangered organisms are in regions characterized as karst, a Modeling; Geochemistry and Climate; Ecosystem Function; term used to describe landscapes containing the caves, sink- Biodiversity; Biological Evolution; Geomicrobiology; and Ap- holes, sinking streams and springs that form in soluble rock, plied Issues, although these titles and topics evolved over the especially limestone. Until fairly recently research on karst in course of the workshop. the U.S. was carried out in a relatively few universities by a few individuals and their students. Research was tightly fo- Following the initial plenary session, participants divided cused on the details of cave ecosystems, karst processes and themselves into seven breakout or focus groups. These initial karst hydrogeology. Also until fairly recently, the karst lands breakout sessions lasted through the end of the first day and of the United States were located in rural areas where their spe- continued through the morning of the second day. In the early cial characteristics caused relatively few problems. Both situ- afternoon of the second day, a second plenary session was con- ations have changed. Research on karst is now underway in vened during which focus group leaders presented the results many institutions, some with substantial groups of faculty and of disciplinary discussions within their group. During this ses- students specializing in the subject. Urbanization and popu- sion, the workshop participants as a whole began to discuss lation growth have pushed into the karst lands so that water research needs that cut across disciplines. The remainder of supplies, ground water contamination, and land hazards such the second day was spent in breakout sessions during which the as sinkhole collapse and sinkhole flooding create increasingly research needs of individual disciplines were refined to reflect large economic and human impacts. Karst research has been the input from the general discussion. The workshop ended on found to have much wider scientific and societal implications the morning of the third day with a final plenary session, where than previously recognized (e.g. paleoclimate archives and wa- there was an open discussion of the findings of the meeting, the ter resources). For the biologists and microbiologists, caves critical research questions that had been identified at the work- have been found to be good models for extreme environments, shop, and how the science should best move forward. a subject of a great deal of current interest. Each focus group leader selected a few experts to participate in The broad array of intellectual challenges facing karst re- their group’s discussion. In addition, general participants could searchers can best be met by multidisciplinary teams who fo- select the focus group closest to their interests and were free to cus their efforts on understanding related problems that cross move between focus groups. Two cross-disciplinary reporters traditional disciplinary boundaries. Such efforts will require were assigned to rove between the groups, transferring ideas extensive planning and highly focused cooperation among dis- and suggestions. parate groups of researchers. The workshop reported in this document had, as its objective, bringing together a representa- The workshop report consists of three parts. In extensive dis- tive cross-section of these scholars so that the current state of cussions following the formal close of the workshop, the or- knowledge could be assessed and some guidance constructed ganizers and the focus group leaders attempted to identify key for the future. areas with the greatest potential for advancement and to distill specific suggestions for future research. These distillations The Karst Waters Institute’s workshop on “Future Directions appear in Part 1 – Opportunities and Recommendations. The in Karst Research” took place in San Antonio, Texas, between focus group leaders were asked to prepare written summaries May 3 and May 6, 2007. Attendees included 86 scientists from of their plenary presentations. These appear as signed reports across North America and Europe. An attempt was made to in Part 2. Focus group leaders also prepared summaries of the include all aspects of karst science including biology, biogeo- discussions within their group which were reviewed and re- chemistry, microbial ecology, hydrogeology and geomorphol- vised by Focus Group members. These appear in Part 3. ogy. The workshop organizers were immensely pleased at the num- The workshop was organized into a series of alternating ple- ber of scientists who took the time and expense to attend. The nary and breakout sessions. The workshop was initiated with large attendance gave the workshop a broad base and an in- a plenary session in which the leaders of each breakout group creased likelihood that the report does indeed represent a cross- presented their views on the current state of knowledge in their section of the scientific community. To all participants, we say discipline. For organizational purposes, the breakout groups Thank You.

Frontiers of Karst Research vii

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13

Part 1

FRONTIERS OF KARST RESEARCH OPPORTUNITIES AND RECOMMENDATIONS: THE NEXT FRONTIER

Opportunities and Recommendations

FRONTIERS OF KARST RESEARCH OPPORTUNITIES AND RECOMMENDATIONS: THE NEXT FRONTIER

arst is the term applied to terrains underlain by soluble Groups are given in Part III. rock: mainly limestone, dolomite and gypsum. Karst Kis a terrain characterized by caves, sinkholes, mobile This summary of research opportunities is organized by scien- soils, sinking streams, fast-flow underground drainage systems, tific context. It does not follow exactly the Focus Group re- large springs, and an assortment of weirdly shaped landforms ports because of overlap and inter-relationships of the subjects sculptured on the bedrock. A recent investigation estimates discussed by the Focus Groups. that karst makes up 12.5% of the Earth’s land surface. AQUIFERS AND GROUND WATER IN Karst impacts human-kind in many ways. Because of the ef- CARBONATE TERRAINS ficient movement of water into the subsurface, surface streams are frequently dry or non-existent. Water wells drilled in karst Opportunities may produce abundant yields or none at all. Karst springs are widely used as water supplies but are at risk of contamination. A reasonable understanding of the components and functioning Soils are easily flushed into the subsurface resulting in “karst of karst aquifers has been established and important progress desertification.” Soil instability can allow foundations tobe has been made on quantitative modeling of aquifer behavior. undermined resulting in extensive property damage. Sinkhole As is frequently the case, present understanding reveals a new collapses are a continuous land-use hazard. layer of scientific questions that would be valuable to investigate. Caves, of many sizes and patterns, are the hallmark features of karst. Caves have been of interest to humans for millennia a) Channel hydraulics: There has been some success at mod- as dwellings, refuges, and places of worship. Only in the past eling flow in conduits in terms of pipes. However, aquifers several centuries have caves become significant objects of sci- may contain reaches of open channel flow interspersed with entific investigation. Caves were found to contain such a fas- reaches of pipe flow. Further, the proportions of each may vary cinating assortment of processes, minerals, unique organisms, with flow volume, flood flows filling more of the system than and sedimentary and paleontological deposits of various kinds base flows. The quantification of this problem in fluid hydrau- that the early investigators invented a new science: speleology. lics would be an important step forward. Speleology was inwardly looking. Scientific disciplines from outside – biology, geology, hydrology, mineralogy, paleontol- b) Turbulent flow: Although difficult to parameterize, tur- ogy, archaeology – were all focused on understanding what was bulent flow processes may explain many observations in karst going on in the caves. flow and transport such as dispersion and tailings in tracer breakthrough curves. Computational fluid dynamics may pro- Greatly accelerated research on caves and karst over the past duce a quantitative model at the laboratory scale but calcula- several decades suggests that the speleologists got it backward. tions at the catchment scale are a more difficult problem. The value of cave and karst related research is not in any press- ing need to understand karst regions as such but for the insights c) Unsaturated flow: Most models and calculations have fo- that karst regions provide to all of the contributing sciences. It cused on aquifer behavior in the phreatic zone. Complete mod- was to identify these promising research opportunities that the els must include the role of the epikarst as a temporary storage Workshop on Frontiers of Karst Science was convened in San and retardation factor and must also include unsaturated flow in Antonio, Texas in early May, 2007. A series of focus group ses- the vadose zone. This would include film flow in wide fractures sions identified subjects for future research that have the poten- and shafts as well as saturated storm flow in smaller fractures. tial for important advances that extend far beyond the research sites themselves. These recommendations are summarized in d) Chemical evolution: What are the mechanisms that control the sections that follow. Disciplinary state-of-the-art keynote the chemical evolution of water along a flow path from surface presentations are given in Part II. The full reports of the Focus water to ground water and return to the surface? This is one of

Frontiers of Karst Research 3

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Opportunities and Recommendations

the best developed aspects of karst but much detail remains to small volume fraction of the aquifer and traditional drilling and be determined. For example, the role of biofilms and microbial geophysical techniques are of limited usefulness. There is need activity in the dissolution and precipitation processes. for novel geophysical approaches that would reveal portions of the conduit system that cannot be reached by explorers. e) High resolution chemical measurements: High resolution temporal and spatial analysis of geochemical and physical Limitations and Concerns variability, particularly for transient events such as storm flows, will be extremely valuable for the understanding and model- The primary limitation is one of commitment. Many previ- ing of the system. Numerous studies have demonstrated that ous hydrologic studies of karst aquifers have been limited to rainfall, changes in water level, spring discharge, turbidity, and relatively short term observations, rarely more than one water concentrations of natural and anthropogenic chemical constitu- year, often much less. To make progress, detailed measure- ents may occur on a timescale of hours. ments need to follow the behavior of benchmark aquifers over many water years. f) Sediment transport: Well-developed karstic aquifers carry a load of clastic sediments which may be stored in the con- KARST AS HUMAN HABITAT duits at base flow and flushed during storm flow. Sediments contribute to the transport of contaminants. The mechanism of Opportunities sediment transport and thresholds for their movement need to be quantified. Living on karst is subject to problems broadly categorized into water supply issues and land use and land management issues. g) Variable density flow: Flow in island and coastal karst is These issues are, in effect, the applied side of karst hydrogeol- modified by the variability in salt concentration. A model for ogy. Aspects in most urgent need of attentions are: the flow is needed that takes account of advective flow and possible density-driven convection. a) Land use impacts: What are the impacts of land use changes and climate change on water resources in karst areas? Can h) Multiphase flow: Primarily, multiphase flow calculations the sources of impact be separated? Such impacts present a would apply to contaminant transport, particularly massive great opportunity for long range quantitative monitoring of dis- spills of non-aqueous phase liquids. However, similar ap- charge and water quality. proaches might be useful for air/water interfaces. The fluid dynamics of such flows in karst aquifers is a challenging prob- b) Contaminant storage: How and when are contaminants lem. stored and later transported through karst aquifers? This will vary with the type of contaminant, the details of aquifer perme- Constraints and Needed Resources ability, and the frequency and intensity of storm flows.

a) Instrumented watersheds: Calculations and model-build- c) Sediment and contaminant transport: What are the re- ing are only as good as their agreement with observation. A lationships between sediment transport and contaminant highly instrumented watershed that would provide detailed transport? The relations between turbidity curves, contami- data on recharge, flow, and discharge over a reasonably long nant chemographs, and hydrographs of karst springs and cave time scale would provide a base against which models could streams in response to storm flows of various magnitudes would be tested. contribute a great deal to this question.

b) Dispersed recharge: Methods need to be developed, in- d) Aquifer yields: Determining sustainability and specific cluding hydrological, geophysical, and remote sensing tech- yield of karst aquifers is, if possible at all, difficult and expen- niques for measuring recharge through the soil zone and the sive. Development of new methods would greatly enhance re- underlying epikarst. Quantitative data on flow and storage source allocation and water supply planning in karst regions. through dry, wet, and storm periods are also needed. e) Sinkholes and land instability: Although there are hun- c) Characterization of the conduit system: Although the con- dreds of case studies documented, there are few fundamental duit system dominates the flow pattern, it is extremely difficult relationships between soil properties, drainage basin character- to detect by other than direct exploration. Conduits comprise a istics and the risk of sinkhole collapse. Sinkhole theory of suf-

4 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Opportunities and Recommendations

ficient accuracy to be an effective prediction tool would be of the transport mechanisms for the sediments, would give insight immense value to land managers. into the hydrologic conditions in existence at the time of deposition. Clastic sediment dates have a low time resolution Constraints and Needed Resources but a much longer time scale – extending into the Pliocene and record rare but major climatic events. Most of the specific research opportunities require long term observations. To accumulate the necessary data a network of b) Relationship between climate change and water avail- dedicated research field sites could be instrumented and man- ability: Calcite growth in speleothems is very dependent on aged for systematic data collection. drip rate from cave ceilings and thus on rainfall. Studies to date have been very promising. Given the large spatial and tem- The study of contaminants in karst aquifers is difficult because poral distribution of speleothems, further studies can provide of the very rapid transport of both water and contaminants. Be- unprecedented insight to: cause serious contaminant spills cannot be introduced under • The underlying mechanisms that have modulated past cli- controlled conditions, an alternative is a “Red Team” approach mate variability across multiple time scales from seasonal in which a selected group of researchers stand ready to mobi- to millennial. lize within hours when a serious spill occurs. This approach • The frequency and magnitude of extreme events such as would allow data collection on the leading-edge of contaminant droughts, floods, and cyclones. behavior which is otherwise rarely observed. • The implications of past variability on future climate change.

Limitations and Concerns c) Climate records in terrestrial environments: Paleocli- mate reconstructions have relied heavily of marine and ice core Although numerous potential dedicated field sites exist, con- records. Caves are widely distributed in the continental inte- tinuous data collection requires up-to-date instrumentation and riors. Profiles of oxygen and carbon isotopes, color and lumi- staffing to assure that the records are complete and collected nescence banding, and trace element profiles provide a wealth with the appropriate temporal and spatial resolution. of climatic proxies. High resolution U/Th dating provides the time scale. The studies made thus far have been highly pro- There are often legal and access barriers to conducting research ductive but each cave provides only one spatial point. Much on contaminated sites due to regulatory issues and liability con- more work is needed to reveal regional and continental scale cerns by land owners. Formal arrangements would be needed patterns. in advance if the “Red Team” approach is to work. d) Karst and the global carbon cycle: To what extent are

CAVES AS CLIMATE ARCHIVES karst systems a source or sink for CO2? The transport of carbon out of soils and into the karst environment and the loss of

Opportunities CO2 from cave atmospheres and karst waters have not been documented on a landscape scale. Karst represents a loop in Cave sediments, both clastic deposits of clays, sands and grav- the carbon cycle that should be addressed. els and speleothems (chemical cave deposits – mostly calcium carbonate) carry imprints of surface conditions at the time of Constraints and Needed Resources their deposition. Combined with techniques for obtaining high accuracy, high temporal resolution dates for these deposits, a) Speleomap: At present, speleothem studies are individual cave sediments are rapidly becoming one of the most important efforts on the part of university faculty and their students. Ac- sources of paleoclimate information on time scales that range cess to the data is only through the published literature which from the historic to the Pleistocene. often does not include the raw data. Significant resources would be required to build a master data base from which a continen- a) Clastic sediments as climate records: Many caves contain tal SPELEOMAP could be produced analogous to the marine deposits of clastic material that can be dated by cosmogenic and lake records used to produce the CLIMAP data set. isotope methods. Such dates provide a chronology for the development of the cave passages and thus the nearby surface b) Calibration: Although speleothems produce outstanding topography. Combining accurate dates with a consideration of high-resolution climate proxies, it is much less certain how to

Frontiers of Karst Research 5

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Opportunities and Recommendations

connect isotope ratios, trace element concentrations, and other b) Origins: What is the source of microbes in caves and time-calibrated data in the speleothem to actual climatic vari- karst? ables on the land surface above the cave. Investigations of the changes in isotope ratios and other parameters along the flow c) Adaptation: Do microbes adapt to the subsurface environ- path between the rainfall and the speleothem drip water are ab- ment? solutely crucial to the interpretation of all other data. d) Habitat: Does the subsurface habitat influence microbial c) Dating: The great value of speleothem records lies in the community composition? ability to determine accurate and absolute dates from milligram samples of the speleothem layers. Only a few laboratories have e) Ecosystem function: How are microbes central to ecosys- the mass spectrometers and know-how to prepare accurate tem function in different types of karst? Are nutrients limiting dates. At present access to dating facilities is through collegial for microorganisms in the subsurface? Are microbial commu- arrangements. Some dating facilities should be supported so nity structure and function linked? that with more formal arrangements they would be open to the entire community. f) Disturbance: Are microorganisms agents or victims of disturbance? Limitations and Concerns Constraints and Needed Resources Paleoclimate investigations of speleothems have the serious drawback that they are destructive. Typically, entire stalag- a) Better tools: Better methods and tools for understanding mites are collected from caves of interest, sectioned, and sam- the presence or absence of microbes are badly needed. These pled along the central core. Speleothems are considered to be should be more inclusive, in situ and non-invasive. Instrumen- a valuable resource and caves have high esthetic value. Strong tation should also be durable, stable, and economical. conservation measures should be respected so that collection is minimized and the amount of data obtained is maximized. b) Standard practices: The application of traditional microbiological and molecular methods may not be appropriate for a) Improved sampling techniques: Microcoring techniques cave and karst investigations. Cultivation studies are needed to should be developed that would permit a slender core to be ex- explore the physiology of indigenous organisms so that better tracted along the growth axis of the stalagmite or other spele- cultivation methods can be developed. othem. It should not be necessary to break off and remove the entire speleothem. Limitations and Concerns

b) Archives: Kilogram quantities of speleothem are removed The limitations are only the usual ones of student interest and to provide milligrams of samples. The unused portions of spe- availability and necessary facilities for research. leothems, along with any characterization data collected concerning them, should be placed in an archive where the material CAVES AND KARST AS HABITATS would be available to other investigators without further damage to the caves. Opportunities

CAVE MICROBIOLOGY Caves act as habitat for limited populations of a limited set of organisms thus making caves useful for the study of ecosystem Opportunities function and species diversity.

a) Microbial diversity. All three domains of life, Eukarya, a) What limits productivity in karst? Microbes are key me- Bacteria, and Archaea, and also viruses occur as microscopic diators of energy flux in food webs that are exclusively hetero- life in caves. There are vast gaps in our knowledge of the diver- grophic as well as those fueled by chemoautotrophy. Explo- sity and distribution of these organisms that needs to be filled. ration of factors limiting microbial productivity would be of Both geographical variation in cave location and also the sub- value. The role of inorganic nutrients in regulating productiv- strate – the character of the mineral surfaces – are important. ity in chemoautotrophic system should be an important avenue of research.

6 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Opportunities and Recommendations

b) Spatial and temporal variation in ecosystem function: CAVES AS EVOLUTIONARY Future research should focus on quantifying the spatial and LABORATORIES temporal patterns in the delivery of energy and the factors that control these patterns. In particular, identifying the presence of Opportunities “hotspots” and “hot times” for ecosystem function and the factors that lead to these situations should be addressed. Cave animals are emerging as strong model systems for understanding the tempo and mode of evolution. Advances in c) The problem of cryptic species: Cryptic species (geneti- molecular and genomic techniques now make it possible to use cally distinct, but morphologically identical) are numerous in these organisms to understand general questions of evolution- caves. The mismatch between genetic convergence and mor- ary biology phological convergence is not understood. What is needed is the simultaneous mapping of genetic and morphological diver- a) How can we understand the relationship between con- sity on a geographic scale. vergence and divergence, and the interplay between morphology and genetics leading to these two paths? While d) Mapping Subterranean Biodiversity: Current data on convergence among disparate taxa is a widespread evolution- subterranean biodiversity is limited by variations in sampling ary phenomenon, the interactions between convergent and di- intensity and frequency, by sampling bias, and by sampling in- vergent phenotypes and genotypes and how these interactions completeness. A project is proposed that would sample 250 affect morphologies have not been comprehensively investi- caves in North America and 250 caves in Europe using stan- gated. Focusing research on particularly obvious convergent dardized sampling protocols. morphological traits (e.g. depigmentation) and the evolution of associated homologous genes across taxonomies will help Constraints and Needed Resources our understanding of general trait evolution, as well as the connection between sequence evolution and protein structural con- A needed resource is a selection of karst drainage basins which straints. can be used for long term study. These need to be controlled and instrumented so that observations can be extended over b) To what extent is the evolution of the cave form con- long time periods. Basins (or portions of basins) where the trolled by the nature of the karst environment versus intrin- land surface can be manipulated by such means as changing sic evolutionary mechanisms? Karst settings offer repeated vegetation, soils and water fluxes would be highly desirable. environments where these types of questions can be investigated. The subterranean habitat exerts strong evolutionary Study of subterranean biodiversity requires the continuing pressures on inhabitants, as evidenced by the particular suite identification of new species. The basic but unglamorous task of traits characterized by highly cave-adapted species. Cave- is limited by the number of trained taxonomists specializing in adapted animals, particularly those with close surface relatives, the various cave organisms. offer the potential for genome-wide comparisons to determine how much of the cave form is genetically hard-wired versus Limitations and Concerns environmentally driven.

A complete understanding of ecosystem function requires un- c) What is the timescale of evolutionary change? Caves derstanding of both mass and energy fluxes through the entire preserve information that is available in few surface habitats, karst system. To obtain this understanding requires communi- where weathering processes remove the record of past events. cation and cooperation between hydrogeologists, geochemists, Well studied karst systems allow for true hypothesis testing, and biologists who, in an ideal arrangement, are studying the where hypotheses about subterranean colonization, lineage di- same system. versification, speciation, and trait evolution can be generated based on the geology of the system and then tested in the evolu- The limited number (and aging) of taxonomists may require in- tion of the species contained within that system. centives for new researchers to enter the field and also perhaps some change of orientation in their training – training students d) Understanding mutation rates: Mutation rates are essen- to be subterranean specialists rather than group specialists, for tial both to understanding both the divergence times of species example. and the evolution of forms. Cave species offer interesting systems in which it may be possible to address these questions.

Frontiers of Karst Research 7

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Opportunities and Recommendations

Constraints and Needed Resources studies but are also important benchmarks for other investigations. There is need to establish at least two pairs of surface and cave- adapted populations or species as model organisms. Cross-Disciplinary Communication

There is need for detailed cooperation between investigators of a) Direct Communication: Existing meetings and journals evolutionary biology and the hydrogeologists and geochemists seem to work reasonably well. Better efforts need to be made concerned with the physical evolution of cave systems, their to develop common language and understanding so that under- relationship to surface topography, and the isotopic age dating standing is not masked by jargon. This is particularly important of cave sediments and speleothems. at the international level.

Limitations and Concerns b) Education: The “graying” of the karst community is a significant concern. New educational initiatives and enhancement These problems are ripe for investigation. of existing cave and karst programs are needed.

CROSS-DISCIPLINARY OPPORTUNITIES Data Bases

An immediately obvious impression obtained from wonder- Raw data concerning chemical composition of waters and cave ing between the Focus Groups was the extent of overlap and deposits, age dates on speleothems, hydrographs and chemo- mutual dependence between the very diverse scientific areas graphs of karst springs, species identifications and distribu- represented. In a certain sense, everyone needs to draw on the tions, and related data are often of value to investigators other concepts and data of everyone else. than the ones who collected the data. Primary data are rarely published. A central data base with clear-cut rules for access Cross-Disciplinary Research and proper use would be invaluable.

a) Redox sensitive elements: What is the fate and transport of Of greater concern are data bases for the caves themselves. The redox-sensitive elements and their microbial consequences and primary data for caves, their locations, descriptions, and maps, feedbacks? Understanding the geochemistry of karst waters re- are collected by private groups of cave explorers. The data quires understanding variable valence (redox) elements such as are held in highly proprietary data bases and there is usually a iron, manganese, and many trace metals. Such understanding great deal of concern that the data will be publically released. A requires also understanding the role of microorganisms in cata- productive working relationship between the karst researchers lyzing and mediating reactions. Cooperative research between and the cavers is essential. geochemists and microbiologists is essential. Archives b) Integration of geochemical data with hydrological modeling: Modeling geochemical processes in conjunction with Much of the potential research requires collecting samples fluid flow is a frontier in hydrologic modeling and would be from caves. This is especially true for paleoclimate studies particularly useful when considering karst dynamics. where entire stalagmites are collected, sectioned, and sampled for dating and isotope analyses. Speleothems are a limited re- c) Epikarst and the vadose zone: Investigation of travel source and there are strong ethical and conservation reasons times and flow paths through the epikarst and the vadose zone for doing as little damage to caves as possible. An archive or is important to hydrogeology, to contaminant transport, to pa- repository for cave material, including any previously collected leoclimate records in caves, and to ecosystem biology. data on chemical, isotopic or petrologic character would allow collected material to be reused, thus minimizing the need for d) Cave development and dating of cave deposits: The time further collections. sequence of cave passage development provides a benchmark for evolutionary biology. Absolute dates on cave deposits, both The archiving of biological and microbiological specimens clastic sediments and speleothems, are critical to paleoclimate should also be considered.

8 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Opportunities and Recommendations

Monitoring Sites, Field Stations, and Field sponse. Such data can only be acquired if funding priority is Laboratories given to establishing intensive monitoring networks in karst systems. NSF’s proposed Hydrologic Observatories would Responses in a karst system can be long term and short term. help achieve this goal. Needed in parallel would be monitor- Long-term (years to centuries) and short-term (hours to weeks) ing networks that cover a maximum of aquifer behavior and time-series data are useful for understanding the varied re- climatic settings.

Frontiers of Karst Research 9

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13

Part 2

TODAY’S FRONTIERS TODAY’S FRONTIERS

It is difficult to plan for the future without knowing where you sive reviews but rather a summary of the highlights. These re- are at the present. All aspects of karst science have made great ports were presented to the entire workshop and were intended strides in the past few decades. It is from this wealth of new to act as a launch point for discussion. The Focus Group lead- knowledge that the workshop participants were able to estimate ers also prepared written versions of their remarks. These are the knowledge needs of the future. reproduced in the pages that follow.

Each Focus Group leader was asked to prepare a short report on These summary papers contain extensive bibliographies which today’s state of knowledge concerning karst in the subject area should guide the reader to the very extensive primary litera- of the Focus Group. These were not intended to be comprehen- ture. Modelling Karst Hydrodynamics

Modelling Karst Hydrodynamics

Attila Kovács and Martin Sauter Geowissenschaftliches Zentrum Universität Göttingen Goldschmidtstrasse 3 37077 Göttingen, Germany

[This article is a lightly edited version the article of the same title that appeared as Chapter 10 in Methods in Karst Hydrology, Nico Goldscheider and David Drew, Eds., Taylor and Francis, London, p. 201-222 (2007). It is reproduced here with the kind permission of Taylor and Francis, Publishers.]

INTRODUCTION The third problem is related to the transfer of the simulated results to a real system. Simplifying assumptions made in the athematical models are exact tools for the quanti- conceptual and numerical models must appear as uncertainties tative representation of the hydraulic behaviour of in the simulation results. Because of the high degree of het- Maquifer systems. The reconstruction of a groundwa- erogeneity, these uncertainties are much larger in karst systems ter flow field, which is consistent with a given hydraulic con- than in most porous unconsolidated aquifers. ductivity field and given boundary conditions, nearly always requires the use of numerical models (Király 2002). Analytical This state-of-the-art report aims to provide a brief overview on models of groundwater flow have been developed since the late karst modelling techniques and related problems. It is intended 1800s. Numerical groundwater flow models have been applied to assist the modelling hydrogeologist in the selection of appro- since the 1960s, and their utilisation for flow through a porous priate tools as well as the definition of suitable parameterisation medium has become every-day practice since then. However, techniques. the application of numerical methods in karst hydrogeology de- mands a specially adapted modelling methodology (Palmer et CONCEPTUAL MODELS OF KARST al. 1999). This is because of particular complexities associated SYSTEMS with the large heterogeneity of a flow field (Király 1994). Every sound conceptual model of karst systems incorporates The first step in any modelling study is the schematic repre- heterogeneity and accordingly the duality of hydraulic flow sentation of a real system. A conceptual model consists of the processes. These include the duality of infiltration, storage, applied differential equations, aquifer geometry, and of a set groundwater flow, and discharge processes. Continuum flow of flow parameters, boundary conditions and initial conditions. (often termed diffuse flow) processes are active in low perme- The hydraulic parameter fields applied in groundwater flow ability matrix blocks or slightly fissured limestone beds, while models are usually obtained with an interpolation between dis- concentrated flow processes can be observed in a discrete con- crete observations. Because of the large heterogeneity and high duit network. Moreover, most conceptual models distinguish contrast in hydraulic conductivity, interpolation techniques three main zones in the vertical direction. These are: soil zone cannot be employed for the characterization of karst aquifers. and epikarst, unsaturated or vadose zone, and phreatic zone. The most demanding task in karst modelling is the definition of Although most conceptual models include similar structural continuous hydraulic parameter fields. features, the flow and storage processes assigned to them display large variations. The second problem is the selection and the development of an appropriate computer code based on numerical methods, which According to the conceptual model of Mangin (1975), the main allows the equations defined in the abstract scheme to be solved. conduit system transmits infiltration waters towards a karst Large heterogeneities require the introduction of special fea- spring, but is poorly connected to large voids in the adjacent tures into the numerical models, such as the combination of rocks, referred to as the ‘annex-to-drain system’. Mangin discrete 1-D elements with a three-dimensional continuum or (1975) associates storage to the ‘annex-to-drain system’ and the double continuum representation of the flow medium. also introduced the concept of epikarst, i.e. a shallow, high-

Frontiers of Karst Research 13

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Modelling Karst Hydrodynamics

permeability karstified zone just below the soil zone. It is be- cally organised karst channel network with a low-permeability lieved to act as a temporary storage and distribution system for fissured rock matrix. While Mangin (1975) associates- stor infiltrating water, similar to a perched aquifer. It is assumed age to the ‘annex-to-drain system’, Drogue (1980) and Király to channel infiltrating water toward enlarged vertical shafts, (1975) associate it to the low-permeability matrix. The strength thereby enhancing concentrated infiltration. of Király’s (1975) conceptual model is based on the fact that it has been tested quantitatively and verified by numerical models Drogue (1974, 1980) assumed, that the geometric configura- (Király and Morel 1976 a, b, Király et al. 1995). A schematic tion of karst conduit networks follow original rock fracture representation of this concept is provided by Doerfliger and patterns. Joints constitute a double-fissured porosity system. Zwahlen (1995) in Figure 1. This network consists of fissured blocks with a size in the order of several hundred metres, separated by high-permeability low MODELLING APPROACHES storage conduits. Every block is dissected by small scale fissures or fractures with considerably larger storage, with low Two, fundamentally different modeling approaches exist for bulk permeability. studying and characterizing karst hydrogeological systems.

The conceptual model proposed by Király (1975, 2002) and Global models (lumped parameter models) imply the math- Király et al. (1995) combines the conceptual models of Mangin ematical analysis of spring discharge time series (hydrographs) (1975) and Drogue (1980). This model involves the epikarst that are believed to reflect the overall (global) hydrogeological and a hierarchical organization of the conduit networks. It also response of karst aquifers. According to this approach, karst comprises the hydraulic effect of nested groups of different systems can be considered as transducers that transform input discontinuities. According to Király and Morel (1976a), two signals (recharge) into output signals (discharge). With the classes of hydraulic parameters can adequately reflect the hy- acquisition of spring discharge data being relatively simple, draulic behaviour of karst systems. Carbonate aquifers can be these models have already been employed since the beginning considered as interactive units of a high-conductivity hierarchi- of the last century. Traditional global modelling techniques do

Figure 1: Conceptual model of karst aquifers (Doerfliger and Zwahlen 1995).

14 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Modelling Karst Hydrodynamics

not take into account the spatial variations within the aquifer Time-series analyses provide limited information concerning and cannot provide direct information on the aquifer geometry, the physical properties of the system itself. Consequently, they hydraulic parameter fields nor physical parameters. However, are “black-box models”. On the other hand, the interpretation some of these techniques combined with other exploration of the results of such models can sometimes be related to some methods may be used for the estimation of hydraulic param- geometric or hydraulic features of karst systems. eters and conduit spacing, information required for distributive flow models. Single Event Methods (Grey Box Models)

For the quantitative spatial simulation of groundwater flow An infiltration event over a karst terrain results in a hydrograph fields, distributive models are needed. The application of dis- peak at the karst spring, which is delayed in time relative to the tributed parameter methods requires the subdivision of a model storm event. This peak can generally be subdivided into three domain into homogeneous sub-units, for which groundwater main components: rising limb, flood recession and baseflow flow can be described by flow equations derived from basic recession (Figure 2). The latter is the most stable section of physical laws. Distributive models can consider both spatial any hydrograph, and also the most characteristic feature of the and temporal variations of hydraulic parameters and boundary global response of an aquifer. This is because baseflow reces- conditions, and thus require detailed information on aquifer ge- sion can be assumed as that segment of a hydrograph least in- ometry, hydraulic parameter fields, and recharge conditions. fluenced by the temporal and spatial variations of infiltration.

GLOBAL METHODS

The measurements of discharge with time at the outlet of a karst aquifer provides integral information on the hydraulic behavior of the entire system. The following two types of spring hydrograph analytical methods can be distinguished (Jeannin and Sauter [1998]).

Single Event Methods deal with the global hydraulic response of the aquifer to a single storm event. It is widely accepted, that Figure 2: Typical features of a spring hydrograph peak. White the hydraulic effect of three fundamental factors are reflected dots indicate inflexion points that correspond with maximum in the global response of a karst aquifer: Recharge, storage infiltration and the end of an infiltration event (Király 1998a, from Kovács 2003). and transmission (flow). Most of the existing techniques for the analysis of spring hydrographs allow the identification of integral parameters of karst properties in a qualitative but not The mathematical description of baseflow recession provided quantitative sense. However, most of these methods are based by Maillet (1905) is based on the depletion of a reservoir, and on simple, or sometimes more complex, cascades of reser- assumes that spring discharge is a function of the volume of voirs and involve physical phenomena. Furthermore, some of water in storage. This behaviour can be described by the fol- these methods provide semi-quantitative relationships between lowing exponential equation: the pattern of the global hydraulic response and hydraulic parameters as well as some geometric properties of an aquifer. Therefore, it is more appropriate to speak of “grey box models” rather than “black-box models”. 3 -1 where Qt is discharge [L T ] at time t, Q0 is the initial discharge Time Series Analyses relate the global hydraulic response of [L3T-1] at an earlier time, and α is the recession coefficient [T-1] karst systems to a succession of recharge events. Univariate usually expressed in 1/day. Plotting a discharge hydrograph on time series analytical methods can identify cyclic variations. a semi-logarithmic graph reveals a straight line with the slope Bivariate time series analyses are designed for the analysis of -a for favourable conditions. This equation is adequate for the the relationship between input (recharge) and output (discharge) description of the baseflow recession of karst systems. parameters for different karst systems. These methods are based solely on mathematical operations, and the coefficients According to Kovács (2003) and Kovács et al. (2005), the obtained cannot directly be related to physical phenomena. baseflow recession coefficient can be used for the derivation of

Frontiers of Karst Research 15

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Modelling Karst Hydrodynamics

important information about aquifer hydraulic parameters and The baseflow recession of fissured systems and poorly karsti- conduit network characteristics. The analytical formulae pro- fied systems is influenced by the hydraulic parameters of frac- vided by the above authors are based on a simple conceptual tures/conduits, low-permeability blocks, fracture spacing, and model, which involves a regular network of high permeability aquifer extent. This flow condition has been defined as con- conduits coupled hydraulically with a low-permeability matrix duit-influenced flow regime (CIFR). The baseflow recession of (Figure 3). The recession coefficient of a karst spring discharg- fractured systems can be expressed as follows: ing from such a system reflects the structure and hydraulic parameters of the conduit network and the low permeability matrix.

The importance of these hydrograph analytical formulae lies in their potential for providing crucial input parameters necessary for distributive groundwater flow models. They also reveal some fundamental characteristics of the recession process of strongly heterogeneous hydrogeological systems, and make the estimation of aquifer parameters and conduit spacing possible from spring hydrograph data.

Forkasiewitz and Paloc (1967) extended Maillet’s approach to the entire recession process by decomposing hydrograph peak recession limbs into three exponential components (Figure 6). The authors assumed that the components reflect three individual reservoirs, representing a conduit network, an intermediate system of well-integrated karstified fractures, and a low per- Figure 3: Conceptual model of karst systems according to meability network of pores and narrow fissures. However, the Kovács (2003) and Kovács et al. (2005). Characteristic model parameters are: Transmissivity of the low-permeability matrix analysis of spring hydrographs simulated by numerical models 2 -1 (Tm [L T ]), storage coefficient of low-permeability matrix (Sm performed by Király and Morel (1976b), and later by Eisenlohr -1 3 -1 [L ]), conductance of karst conduits or fractures (Kc [L T ]), -1 et al. (1997a) showed that different exponential hydrograph storage coefficient of karst conduits (Sc [L ]), spatial extent of the aquifer (A [L2]), and the spatial frequency of karst conduits segments do not necessarily correspond to aquifer volumes (f [L-1]). with different hydraulic conductivities. Three exponential reservoirs can be fitted to the hydrograph of a karst system consisting of only two classes of hydraulic conductivities. The Numerical studies by the above authors showed that the depen- intermediate exponential could simply be the result of transient dence of the recession coefficient on aquifer properties cannot phenomena in the vicinity of the high hydraulic conductivity be described using a single formula, but that it follows two sig- channel network. nificantly different physical principles depending on the overall configuration of the hydraulic and geometric parameter fields Other authors proposed to describe the recession process by (Figure 4, Figure 5). employing different mathematical functions. Drogue (1972) described the whole recession process by using one single hy- The baseflow recession of karst systems is controlled by the hy- perbolic formula. In contrast, Mangin (1975) distinguished draulic parameters of the low-permeability matrix, and conduit two components on the recession curves, and associated them spacing. This flow condition is referred to as matrix-restrained to the discharge from the unsaturated zone and the saturated flow regime (MRFR). The baseflow recession coefficient of zone, respectively. karst systems can be expressed as follows: Time Series Analysis

Time series analyses imply the mathematical analysis of the response of karst systems to a succession of rainfall events. Because these methods are solely based on mathematical op-

16 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Modelling Karst Hydrodynamics

erations, they fail to provide information on physical functioning of karst systems, and can be used mainly for prediction and data compilation purposes. Most of the methods used in time series analysis were principally developed by Jenkins and Watts (1968), and were applied to karst systems by Mangin (1971, 1975, 1981, 1984). Detailed explanation of time-series analytical techniques is provided by Jeannin and Sauter (1998).

Figure 4: Graphical representation of the relationship between recession coefficient and conduit conductance on log-log graph. Kc* represents the threshold value of conduit conductance. Higher values entail matrix-restrained baseflow (MRFR), while lower values result in conduit-influenced baseflow (CIFR). Modified after Kovács (2003) and Kovácset al. (2005) Figure 6: Decomposition of recession curves according to Forkasiewicz and Paloc (1967).

Conventional time series analysis uses both univariate (autocorrelation, spectral analysis) and bivariate (cross-correlation, cross-spectral analysis) methods. Univariate methods charac- terise the structure of an individual time series (hydrograph), while bivariate methods describe the transformation of an input function into an output function (rainfall/discharge).

The Auto-Correlation Method is a tool for the identification of some overall characteristics of a discharge time series, particularly cyclic variations (Mangin 1982, Grasso and Jeannin, 1994, Eisenlohr et al., 1997b). Spectral analysis offers considerable potential as a powerful tool for the analysis of periodicities within the time series (Box and Jenkins 1976, Mangin 1984).

Cross correlation methods permit the quantitative comparison of rainfall with spring discharge time series. The technique provides information about strength of the relationship between the two time series and the time lag between them (Jenkins and Figure 5: Graphical representation of the relationship between Watts 1968, Box and Jenkins 1976, Mangin 1981, 1982, 1984, recession coefficient and conduit frequency on log-log graph. Padilla and Pulido-Bosch 1995, Larocque et al., 1998, Grasso f* represents the threshold value of conduit frequency. Higher 1998, Grasso and Jeannin 1998). values entail conduit-influenced baseflow (CIFR), while lower values result in matrix-restrained baseflow (MRFR). Modified after Kovács (2003) and Kovács et al. (2005). In the case of linear and steady state systems, a transfer function can be defined, which is a characteristic function of the

Frontiers of Karst Research 17

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Modelling Karst Hydrodynamics

system that reflects the active processes. The method of iden- where S is the storage coefficient [L-1], K is hydraulic conduc- tifying the transfer function of an unknown system is known tivity [LT-1], H is hydraulic head [L], t is time [T], and i is the as deconvolution. Assuming the input function to be random, source term [T-1]. the cross-correlogram can be considered as the transfer function of the system (Neuman and De Marsily 1976, Dreiss 1989, Two types of discretisation methods are used widely in hy- Mangin 1984). drogeology. The Finite Difference Method (FDM) consists of subdividing the model domain into rectangular cells. Partial Discharge time series may exhibit non-stationarity in their sta- derivatives are then approximated by simple differences be- tistics. While the series may contain dominant periodic signals, tween a given number of adjacent nodes located at the corners these signals can vary both in amplitude and frequency over or in the centre of each cell. According to the Finite Element long periods of time. One possibility for the analysis of non- Method (FEM), the model domain is subdivided into an irregu- stationarity of a time series would be to compute a Windowed lar network of triangular and/or quadrangular finite elements. Fourier Transform using a certain window size, and sliding it The approximation of differential operators is analytical, and it along the series. This would provide information on frequency involves integral quantities for each element. The FEM allows localization, but would still be dependent on the window size the combination of one-, two- and three-dimensional elements used, resulting in an inconsistent treatment of different frequen- of various shapes, thus facilitating proper discretisation and the cies (Torrence and Compo 1998). implementation of discrete features into the model. A detailed explanation of the FDM and FEM discretisation methods can Wavelet analysis introduced by Grosmann and Morlet (1984) be found in Kinzelbach (1986), Wang and Anderson (1982), attempts to solve this problem by decomposing a time series and Huyakorn and Pinder (1983). into time/frequency space simultaneously. The continuous wavelet transform is defined as the convolution of the time se- The equation system obtained by spatial discretisation requires ries with a scaled and translated version of a wavelet function matrix inversion. This can be easily computerised, and thus (Torrence and Compo 1998). Several types of wavelet func- spatial discretisation allows groundwater flow simulations in tions are in use; each must have zero mean and be localised in complex hydrogeological systems. The solution of an equa- both time and frequency space. tion system requires the definition of boundary conditions and (for transient problems) initial conditions. Boundary condi- Wavelet analysis constitutes an alternative in karst hydrology to tions consist of the definition of either hydraulic potential (head spectral and correlation analyses. By varying wavelet scale and boundary) or flux (flux boundary) values along the domain sliding the wavelet along a time series, one can construct a pic- boundaries. Initial conditions consist of the spatial distribution ture showing both the amplitude of any periodic signals versus of the unknown function, and are usually taken from an equipo- scale and the variation of this amplitude with time. A theoreti- tential map or a previous steady state simulation. cal explanation of wavelet analysis is provided by Daubechies (1992). The application of wavelets in karst hydrogeology is Distributed parameter groundwater flow models include two discussed in Labat et al. (1999 a, b, 2000). principal concepts. The discrete concept considers the flow within individual fractures or conduits. In contrast, the contin- DISTRIBUTIVE METHODS uum concept treats heterogeneities in terms of effective model parameters and their spatial distribution. These concepts can be The spatial heterogeneity of karst aquifers may require the dis- combined into five alternative modelling approaches, according cretisation of a hydrogeological system into homogeneous sub- to the geometric nature of the conductive features represented units, each with its own characteristic hydraulic parameters. in the model (Teutsch and Sauter 1991, 1998) (Figure 7): The discretisation of a rock volume involves the discretisation of the differential equations describing groundwater flow. The • Discrete Fracture Network Approach (DFN) principal formula representing transient groundwater flow in • Discrete Channel Network Approach (DCN) saturated medium is the classical diffusivity equation derived • Equivalent Porous Medium Approach (EPM) from Darcy’s flow law (momentum conservation) and the con- • Double Continuum Approach (DC) tinuity equation (mass conservation): • Combined Discrete-Continuum (Hybrid) Approach (CDC)

18 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Modelling Karst Hydrodynamics

The physical parameters of the flow medium can be directly nificantly underestimates the storage and overestimates veloc- or indirectly derived from real field observations (deterministic ity. For this reason, it is preferable to derive fracture properties models) or can be determined as random variables (stochastic directly from field experiments. Packer tests provide informa- models). Each method has its respective advantages and limi- tion on transmissivity, interference tests estimate storage, and tations, and the selection of the appropriate modelling approach tracer tests determine transport aperture. may be crucial with respect to the outcome of the simulation.

Figure 7: Classification of distributive karst modelling methods.

Discrete Fracture Network Approach (DFN)

According to the DFN approach, only certain sets of fractures are considered to be permeable. The matrix medium is assumed to have negligible permeability. This concept simplifies a fissured system into a network of two-dimensional fracture Figure 8: Example of a DFN network consisting of two fracture planes (Figure 8). It is mainly applicable to fractured aquifers. sets. However, DFN methods facilitate the representation of karst channels by introducing one-dimensional elements or increased The first two-dimensional DFN modelling of synthetic fracture transmissivity zones along individual fractures representing networks using a numerical solver was published by Long et al. dissolution voids (Dershowitz et al., 2004). (1982). Fracture networks were generated stochastically by applying a Monte Carlo process. Long et al. (1985) constructed a The transmissivity T [L2T-1] of a single fracture can be ex- three-dimensional DFN model, and simulated steady-state flow pressed by the “cubic law” (Snow 1965) as follows: for simple theoretical configurations of orthogonal fractures. Andersson and Dverstorp (1987) presented a modelling case study for conditioning the statistical generation of fracture sets with field data.

According to the Monte Carlo process, probability density where a is the fracture aperture [L], μ is the dynamic viscosity functions are fitted to field data describing the spacing, length, of water [ML-1T-1], ρ is fluid density [ML-3], and g is accelera- direction and aperture of fracture sets identified on stereonets. tion due to gravity [LT-2]. The “cubic law” is valid for laminar Parameters are sampled from these statistical distributions, and flow in open or closed fractures (Witherspoon et al., 1980). assigned to individual fractures in the model. Fracture plane The storativity of a single fracture assuming water compres- centres are generated first, assuming a negative exponential sion only, can be expressed as follows: distribution of fracture spacing (Baecher et al., 1977). Fracture sizes are then assigned and usually sampled from lognormal distribution functions. Fracture orientations usually show a Fisher distribution, while fracture transmissivities follow a lognormal distribution.

A major benefit of DFN models is their ability to reflect the The cubic law is based on a parabolic distribution of velocities compartmentalization phenomenon experienced at several between smooth parallel plates, which is a condition seldom study sites (Shapiro and Hsieh 1994, Sawada et al., 2000). met in natural fractures. Natural fractures are rough, tortuous Compartments are adjacent rock volumes that manifest signifi- and heterogeneous. The cubic law aperture can therefore sig- cant differences in hydraulic heads (up to 100 meters difference

Frontiers of Karst Research 19

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Modelling Karst Hydrodynamics

on a few meters of distance). Robinson (1984) demonstrated that there was a critical number of intersections required to produce percolating pathways. Compartments are the result of insufficient density of the fracture network. The porous medi- where K’ is turbulent flow effective hydraulic conductivity um concept assumes hydraulic continuity between all points in [LT-1], A is cross sectional area [L2], and I is hydraulic gradient the simulation region, and cannot reflect compartmentalization [-]. Louis (1968) expressed the effective hydraulic conductiv- phenomenon. ity for fully constricted (phreatic) pipe flow as follows:

One of the most critical shortcomings of the DFN approach is its high data demand. While fracture spacing, orientation and transmissivity distributions can be estimated from downhole measurements, fracture size parameters usually remain uncer-

tain. Computational constraints limit the number of simulated where Dh=4Rh is the hydraulic diameter [L], and ε is the abso- fractures to the order of 104 to 105. Fracture size distributions lute size of irregularities along conduit walls [L]. Similarly, are usually truncated, and small size fractures are not repre- Strickler (1923) expressed the effective hydraulic conductivity sented in the model. This results in an erroneous estimation of for non-constricted (vadose) pipe flow as follows: the storage and diffusion behaviour, and necessitates the implementation of artificial matrix blocks in the model. As fracture parameters result from a stochastic generation method, model results involve significant spatial uncertainties.

1/3 -1 Discrete Conduit Network Approach (DCN) where Ks is the Strickler coefficient [L T ] depending on

roughness. Rh is the hydraulic radius (cross section divided by The DCN approach simulates flow in networks of one-dimen- wet perimeter) [L]. The storage coefficient of a conduit as- sional pipes representing karst conduits or connections between suming water compression only, can be formulated as follows fracture centres. A DCN conduit network can be established (Cornaton and Perrochet, 2002): deterministically representing a local-scale field situation, or can be derived from stochastic DFN networks by geometric transformations.

The mathematical formulation of the average velocity of laminar flow in one-dimensional conduits may be expressed by the A summary and comparison of these formulae, along with Hagen-Poiseuille law: case studies for their application is provided by Jeannin and Maréchal (1995), and Jeannin (2001). These authors constructed a two-dimensional fully constricted pipe-flow model of the Hölloch Cave, Muotatal, Switzerland, in order to simulate groundwater flow for various discharge conditions (Figure 9). The geometric properties of the conduits were derived from field observations. The model was calibrated on the basis of where r is conduit radius [L]. Conduit conductance is a one- several head and discharge measurements in conduits. This dimensional parameter derived from the Poiseuille law as fol- study demonstrated how overflow conduits modify aquifer lows: conductivity when they become active. This results in the observed large non-linearity of the system. The typical effective hydraulic conductivity of karst conduits ranges between 1 and 10 m/s and the Louis formula is adequate for calculating head- losses under such conditions.

Another application of the DCN approach is for the representa- The mathematical formulation of turbulent flow in one-dimen- tion of fracture networks (Cacas et al., 1990, Dverstorp et al., sional conduits is given by the Darcy-Weissbach friction law: 1992, Dershowitz et al., 1998). Fractures are usually represent-

20 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Modelling Karst Hydrodynamics

Figure 9: The DCN model of the downstream part of the Hölloch cave, Muotatal, Switzerland (Jeannin 2001). The parameter k’s is equivalent to K’A.

ed by one-dimensional elements connecting fracture intersec- the fracture properties contained in that cell. The result is an tions. DCN representation of fracture networks may be useful equivalent permeability tensor, according to a specific grid. when no spatial information on fracture flow is necessary. DCN Fine discretisation reproduces the hydraulic and transport be- networks derived from DFN models can correctly represent haviour of the underlying fractured medium. However, for the overall transport behaviour of fissured systems, and can be coarser practical discretisations of tens or hundreds of meters, used for simulating breakthrough curves. The transformation the Oda approximation produces less accurate results. of fractures into one-dimensional pipes requires the estimation of effective pipe width, which remains a calibration parameter. The differences in transient hydraulic behaviour between fis- The quality of DCN models in this case strongly depends on the sured, karstic and equivalent porous systems were demon- quality of the DFN model from which they originate. strated by Kovács et al. (2005). These authors demonstrated that porous systems manifest fundamentally different temporal Equivalent Porous Medium Approach (EPM) hydraulic behaviour from karstified medium. The global response of EPM models corresponds to the transition between The EPM approach utilizes discretization units of similar size. matrix-restrained and conduit-influenced flow (Figure 10). This requires the representative elementary volume to be almost constant all over the model domain, and involves an in- The application of EPM models for steady-state karst hydro- significant change of aquifer hydraulic parameters between geological problems is mainly limited by the constraints of adjacent units of discretization. This is a condition seldom met discretisation. Transient EPM models fail to reflect karstic hy- in karst systems. draulic behaviour. As a consequence, the EPM approach is basically inadequate for modelling karst hydrogeological systems The substitution of strongly heterogeneous rocks with equiva- over a large spectrum of applications. lent porous medium (this transformation referred to as upscal- ing) has long been an interest of hydrogeologists active in the Double Continuum Approach (DC) field of fractured rock hydrology. Oda (1985) presented a method for calculating EPM properties from fracture networks. The difficulties of obtaining data for constructing DCN or CDC The Oda approach overlays an EPM grid across a fracture net- models, and the inability of the EPM approach to take account work, and derives EPM properties for each grid cell based on of the strong heterogeneity of karst aquifers, have motivated

Frontiers of Karst Research 21

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Modelling Karst Hydrodynamics

The α parameter [L-1T-1] characterises the rate of the fluid transfer between the two media, which is composed of the hydraulic conductivity of the exchange zone and geometric factors. The flux value of a node (e.g. spring discharge) can then be evaluated as the sum of the fluxes from the two different media, while head values are observed separately in the two media.

Figure 10: Variations of the recession coefficients for equiva- Figure 11. Schematic representation of a one-dimensional lent discrete-continuum models. EPM models correspond DC model (Teutsch 1988). to the transition between matrix-restrained and conduit-influenced flow, and designates an inflection point ofthere- Although the DC concept can adequately describe the dual hy- cession coefficient curve. Modified after Kovács (2003) and draulic behaviour of karst aquifers, applied hydraulic param- Kovács et al. (2005). eter distributions are volume averaged parameters of the real parameters. As the calibration of the parameters is basically the development of the DC method, which can simulate spe- a “trial- and-error” method based on available head data, the cific karst features without requiring detailed knowledge of the quality of the model results strongly depends on the density and conduit network geometry. The first numerical solution using location of the observation points. To some extent the param- the DC approach was provided by Teutsch (1988), who used eters of the DC system can be estimated from pumping tests, the original concept of Barenblatt et al. (1960) (Figure 11). In especially for the matrix system. Sauter (1992) proposed pa- a double continuum model the conduit network and the fissured rameter estimation methods for DC models. medium are both represented by continuum formulations. The exchange of water and solute between the two continua is cal- The adequacy of DC model results and calibrated parameter culated based on the hydraulic head difference between them, fields has been tested by several authors. Mohrlok & Teutsch using a linear exchange term. Flow equations for the two sepa- (1997) demonstrated that simulated spring discharges obtained rate media may be formulated as follows: from calibrated DC models show a good correspondence to field observations. Cornaton (1999) tested the physical mean- ing of the calibration parameter, and demonstrated a strong dependence of the exchange coefficient on matrix storage and conduit network density. The author also provided a one-di- (13) mensional analytical solution for the problem.

Where i is the identifier of the medium considered, the last Another critical aspect of the DC concept is that it cannot handle term is the exchange flux between the two continua, calculated the temporal delay of diffuse infiltration, since both subsystems from the hydraulic head difference, multiplied by an empirical are coupled directly at every node. In order to involve retarded steady-state exchange coefficient (α) as follows: diffuse infiltration in a DC model, a retention function is necessary (Sauter, 1992). Two-dimensional real-world applications of the DC concept were successfully performed by Teutsch (1988), Sauter (1992), and Lang (1995). Three-dimensional

22 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Modelling Karst Hydrodynamics

real-world applications of the DC method have not yet been els of karst systems. As demonstrated by Kovács (2003) and performed because of the difficulties in model calibration and Kovács et al. (2005), the correct estimation of conduit network the necessity of requiring three-dimensional head data. density and hydraulic parameters is crucial for modelling karst systems by the CDC approach. There is only one single pa- The DC approach is an effective tool for modelling karst sys- rameter configuration that yields appropriate transient model tems. The spatial distribution of the calibration parameter val- results. These authors also demonstrated, that epikarstic stor- ues provides approximate information on real conduit system age can significantly influence the global reactions of a karst configuration. However, the interpretation of applied hydraulic system, and thus such systems require the separate estimation parameters requires further attention. of epikarst hydraulic and geometric parameters.

Combined Discrete-Continuum (Hybrid) Approach (CDC)

The CDC approach is capable of handling the discontinuities that exist at all scales in a karst system (fractures, fault zones, karst channels, etc.), by representing them as embedded networks of different orders of magnitude. This nested model ex- plains the duality of karst (duality of the infiltration process, the flow field and discharge conditions), and the scale effect on hydraulic conductivities (Király 1975).

Because the CDC approach uses the FEM discretisation method, it allows the combination of one-, two-, and three-dimensional elements (Király 1979, 1985, 1988, Király et al., 1995, Király and Morel, 1976a). According to the CDC method, high conductivity karst channels can be simulated by one-dimensional finite elements, which are set in the low permeability matrix represented by three-dimensional elements. Similarly, the application of two-dimensional elements makes the simulation of fractures and fault zones possible.

The CDC method was developed and first applied by Király and Morel (1976a). This was the first time that the typical temporal behaviour of a karst spring was adequately simulated by a distributive model, using realistic hydraulic parameter distributions (Figure 12) Király (1998b) concluded that the application Figure 12. Finite element mesh and simulated spring hy- of EPM models requires artificially high hydraulic conductivi- drographs of the first CDC model (Areuse Basin, Switzer- ties in order to simulate realistic hydraulic heads. However, land), Király and Morel (1976a). these discrepancies disappear with the introduction of the high conductivity channel network into the FEM model. Typical spring hydrographs cannot be simulated without applying con- REFERENCES centrated infiltration. The rate of this concentrated input should be more than 40 % of the total infiltration. The epikarst layer Andersson, J. and Dverstorp, B., 1987, Conditional simulations (Mangin 1975) may play an important role in draining and in- of fluid flow in three-dimensional networks of discrete frac- filtrating water, and channelling it towards the karst network. tures: Water Resources Research, v. 23, p. 1876-1886. Baecher, G.B., Lanney, N.A., and Einstein, H.H., 1977, Statis- The CDC approach is the only distributive modelling con- tical descriptions of rock properties and sampling: Proceedings cept that facilitates the direct application of observed aquifer of the 18th U.S. Rock Mechanics Symposium, p. 5C1-5C1-8. geometry and measured hydraulic parameters. Consequently, the CDC approach facilitates the testing of conceptual mod-

Frontiers of Karst Research 23

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Modelling Karst Hydrodynamics

Barrenblatt, G.I., Zheltow, I.P., and Kochina, I.N., 1960, Ba- moires hors série Société Géologique de la France, v. 11 p. 101- sic concepts in the theory of seepage of homogeneous liquids 108. in fissured rocks (strata): Journal of Applied Mathematics and Mechanics, v. 24, p. 1286-1303. Dverstorp, B., Andersson, J., and Nordqvist, W., 1992, Dis- crete fracture network interpretation of field tracer migration Box, G.E.P. and Jenkins, G.M., 1976, Time series analysis: in sparsely fractured rock: Water Resources Research, v. 28, p. forecasting and control. Holden Day, San Francisco. 2327-2343

Cacas, M.C., Ledoux, E., deMarsily, G., Tilie, B., Barbreau, Eisenlohr, L., Király, L., Bouzelboudjen, M., and Rossier, I., A., Durand, E., Feuga, B., and Peaudecerf, P., 1990, Modeling 1997a, A numerical simulation as a tool for checking the inter- fracture flow with a stochastic discrete fracture network model: pretation of karst springs hydrographs: Journal of Hydrology, calibration and validation, 1. the flow model: Water Resources v. 193, p. 306-315. Research, v. 26, p. 479-489. Eisenlohr, L., Király, L., Bouzelboudjen, M., and Rossier, I., Cornaton, F., 1999, Utilisation de modeles continu discret et 1997b, Numerical versus statistical modeling of natural re- a double continuum pour l’analyse des reponses globales de sponse of a karst hydrogeological system: Journal of Hydrol- l’aquifere karstique : Thesis, CHYN, University of Neuchatel. ogy, v. 202, p. 244-262.

Cornaton, F. and Perrochet, P., 2002, Analytical 1D dual-poros- Forkasiewicz, J. and Paloc, H., 1967, Le régime de tarisse- ity equivalent solutions to 3D discrete single-continuum mod- ment de la Foux-de-la-Vis. Etude préliminaire: Chronique els. Application to karstic spring hydrograph modeling: Journal d`Hydrogéologie, BRGM, v. 3(10), p. 61-73. of Hydrology, v. 262, p. 165-176. Grasso, D.A., 1998, Interprétation des réponses hydrau- Daubechies, I., 1992, The wavelet transform time-frequency liques et chimiques des sources karstiques: Thèse, Centre localization and signal analysis. IEEE Transactions on Infor- d’hydrogéologie, Université de Neuchâtel. mation Theory, v. 36, p. 961-1004. Grasso, D.A. and Jeannin, P-Y., 1994, Etude critique des Doerfliger, N. and Zwahlen, F., 1995, Action COST 65 – Swiss méthodes d’analyse de la réponse globale des systèmes National Report: Bulletin d’Hydrogéologie de l’Université de karstiques. Application au site de Bure (JU, Suisse): Bulletin Neuchâtel, v. 14, p. 3-33. d`Hydrogéologie Neuchâtel, v. 13, p. 87-113.

Dershowitz, W.S., Toxford, T., Sudicky, E., Shuttle, D.A., Ei- Grosmann, A.and Morlet, J. 1984, Decomposition of Hardy ben, T.H., and Ahlstrom, E., 1998, PA Works pathways analysis functions into square integrable wavelets of constant shape: for Discrete Fracture Networks with LTG solute transport: User SIAM Journal of Mathematical Analysis, v. 15, p. 723-736. Documentation, Golder Associates Inc., Redmond, Washing- ton. Huyakorn, P.S. and Pinder, G.F., 1983, Computational methods in subsurface flow: Academic Press, London. Dershowitz, W.S., La Pointe, P.R., and Doe, T.W., 2004, Ad- vances in discrete fracture network modeling: Proceedings of Jeannin, P-Y. (2001) Modeling flow in phreatic and epiphreatic the U.S. EPA/NGWA Fractured Rock Conference. Portland, karst conduits in the Hölloch Cave (Muotatal, Switzerland): OR. Water Resources Research, 37(2): 191-200.

Dreiss, S.J., 1989, Regional scale transport in karst aquifer. 2: Jeannin, P-Y. and Maréchal, J-C., 1995, Lois de pertes de Linear systems and time moment analysis: Water Resources charge dans les conduits karstiques: base théorique et observa- Research, v. 25, p. 126-134 tions: Bulletin d’Hydrogéolgie, Neuchatel, v. 14, p. 149-176.

Drogue, C., 1972, Analyse statistique des hydrographes de de- Jeannin, P-Y. and Sauter, M., 1998 Analysis of karst hydrody- crues des sources karstiques: Journal of Hydrology, v. 15, p. namic behaviour using global approaches: a review: Bulletin 49-68. d`Hydrogéologie, Neuchatel, v. 16, p. 31-48.

Drogue, C., 1974, Structure de certains aquifères karstiques Jenkins, G.M. and Watts, D.G., 1968, Spectral analysis and its d’après les résultats de travaux de forage: Comptes Rendus applications. Holden Days, San Francisco. Académie des sciences, Paris, D, v. 278, p. 2621-2624. Kinzelbach, W., 1986, Groundwater modeling. Elsevier, Inter- Drogue, C., 1980, Essai d’identification d’un type de structure national Edition. de magasins carbonates fissurées. Application à l’interprétation de certains aspects du fonctionnement hydrogéologique: Mé-

24 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Modelling Karst Hydrodynamics

Király, L., 1975, Rapport sur l’état actuel des connaissances Labat, D., Ababou, R. and Mangin, A., 1999a, Analyse en on- dans le domaine des caractères physique des roches karstique. dolettes en hydrogeology karstique. 1re partie: analyse univar- In: Burger, A. and Dubertet, L.(eds.), Hydrogeology of karstic iée de pluies et débits de sources karstiques: Earth and Plan- terrains, International Union of Geological. Sciences, ser. B, etary Science, v. 329, p. 873-879. No. 3, p. 53-67 Labat, D., Ababou, R. and Mangin, A., 1999b, Analyse en on- Király, L., 1979, Remarques sur la simulation des failles dolettes en hydrogeology karstique. 2e partie: analyse en on- et du réseau karstique par éléments finis dans les modèles dolettes croisée pluie-debit: Earth and Planetary Science, v. d’écoulement: Bulletin d’Hydrogéologie de l’Université de 329, p. 881-887. Neuchâtel, v. 3, p. 155-167. Labat, D., Ababou, R. and Mangin, A., 2000, Rainfall-runoff Király, L., 1985, FEM-301 – A three dimensional model for relations for karstic springs: continuous wavelet and discrete groundwater flow simulation: NAGRA Technical Report 84- orthogonal multiresolution analyses: Journal of Hydrology, v. 49, 96 p. 238, p. 149-178.

Király, L., 1988, Large-scale 3D groundwater flow modeling in Lang, U., 1995, Simulation regionaler stromungs und trans- highly heterogeneous geologic medium: In Custoido et al. eds., portvorgange in karstaquiferen mit hilfe des doppelkontinuum- Groundwater flow and quality modeling, D. Riedel Publishing ansatzes: Methodenentwicklung und parameteridentifikation: Company, p. 761-775. Ph.D. thesis, University of Stuttgart.

Király, L., 1994, Groundwater flow in fractures rocks: models Larocque, M., Mangin, A., Razack, M., and Banton, O., 1998, and reality: In 14th Mintrop Seminar über Interpretationsstrat- Characterization of the La Rochefoucauld karst aquifer (Char- egien in Exploration und Produktion, Ruhr Universität Bochum ente, France) using correlation and spectral analysis: Bulletin 159, p. 1-21. d`Hydrogéologie (Neuchâtel), v. 16, p. 49-57.

Király, L., 1998a Introduction à l’hydrogéologie des roches fis- Long, J.C.S., Remer, J.S., Wilson, C.R., and Witherspoon, P.A. surées et karstiques. Bases théoriques à l’intention des hydro- (1982) Porous media equivalents for networks of discontinuous géologues: Manuscrit, Université de Neuchâtel. fractures: Water Resources Research, v.18, p. 645-658.

Király, L., 1998b, Modeling karst aquifers by the combined Long, J.C.S, Gilmour, P., and Witherspoon, P.A., 1985, A mod- discrete channel and continuum approach: Bulletin du Centre el for steady fluid flow in random three-dimensional networks d’Hydrogeologie, Neuchatel, v. 16, p. 77-98. of disc-shaped fractures: Water Resources Research, v. 21, p. 1105-1115. Király, L., 2002, Karstification and Groundwater Flow:In Pro- ceedings of the Conference on Evolution of Karst: from Pre- Louis, C., 1968, Etude des écoulements d’eau dans les roches karst to Cessation, Postojna-Ljubljana. p. 155-190. fissurées et de leurs influences sur la stavilité des massifs roch- eux: Bull. Dir. Étud. Rech. Electr. Fr., A3, p. 5-132. Király, L. and Morel, G., 1976a, Etude de régularisation de l’Areuse par modèle mathématique: Bulletin du Centre Maillet, E., 1905, Essais d`hydraulique souterraine et fluviale: d’Hydrogéologie, Neuchâtel, v. 1, p. 19-36. Hermann, Paris.

Király, L. and Morel, G., 1976b, Remarques sur l’hydrogramme Mangin, A., 1971, Etude des débits classés d’exutoires des sources karstiques simule par modèles mathématiques: Bul- karstiques portant sur un cycle hydrologique: Annales de Spé- letin du Centre d’Hydrogéologie, Neuchâtel, v.1, p. 37-60. léologie, v. 28, p. 21-40.

Király, L., Perrochet, P. and Rossier, Y., 1995, Effect of epikarst Mangin, A., 1975, Contribution a l`étude hydrodynamique des on the hydrograph of karst springs: a numerical approach: Bul- aquifères karstiques: Thèse, Institut des Sciences de la Terre de letin d’hydrogeologie, v. 14 p. 199-220. l`Université de Dijon.

Kovács, A., 2003, Geometry and hydraulic parameters of karst Mangin, A., 1981, Utilisation des analyses correlatoire et spec- aquifers: A hydrodynamic modeling approach: Ph.D. Thesis, trale dans l’approche des systemes hydrologiques: Comptes University of Neuchâtel, Switzerland, 131 p. Rendus de l’Académie des Sciences, Série III, v. 293, p. 401- 404 Kovács, A., Perrochet, P., Király, L. and Jeannin, P.-Y., 2005, A quantitative method for the characterization of karst aquifers Mangin, A., 1982, L’approche systémique du karst, con- based on spring hydrograph analysis: Journal of Hydrology, v. séquences conceptuelles et méthodologiques: Proceedings 303, p. 152-164. Réunion Monographica Sobre el Karst, Larra. p.141-157

Frontiers of Karst Research 25

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Modelling Karst Hydrodynamics

Mangin, A., 1984, Pour une meilleure connaissance des sys- Snow, D.T., 1965, A parallel plate model of fractured perme- temes hydrologiques à partir des analyses corrélatoire et spec- able media: PhD Thesis, University of California Berkeley. trale: Journal of Hydrology, v. 67, p. 25-43. Sauter, M., 1992, Quantification and forecasting of regional Mohrlok, U. and Teutsch, G., 1997, Double continuum porous groundwater flow and transport in a karst aquifer (Gallusquelle, equivalent (DCPE) versus discrete modeling in karst terranes: Malm, SW. Germany): Tübinger Geowissenschaftliche Arbe- In Günay, G. and Johnson, K (eds.), Karst waters and environ- iten, C13. mental impacts, Balkema, Rotterdam. Strickler, A., 1923, Beitrage zur Frage der Geschwindigkeits- Neuman, S.P. and De Marsily, G., 1976, Identification of linear formel und der Rauhigkeitszahlen für Ströme, Kanäle und system response by parametric programming: Water Resources geschlossene Leitungen: Mitt. Amt. für Wasserwirt, v. 16, p. Research, v. 12, p. 253-262. 21-38.

Oda, M., 1985, Permeability tensor for discontinuous rock Teutsch, G., 1988, Grundwassermodelle im Karst: Praktische masses: Geotechnique, v. 35, p. 483-495. Ansätze am Beispiel zweier Einzugsgebiete im Tiefen und Se- ichten Malmkarst der Schwäbischen Alb: Ph.D. Thesis, Uni- Padilla, A. and Pulido-Bosch, A., 1995, Study of hydrographs versity of Tubingen. of karstic aquifers by means of correlation and cross-spectral analysis: Journal of Hydrology, v. 168, p. 73-89. Teutsch, G. and Sauter, M., 1991, Groundwater modeling in karst terrains: Scale effects, data acquisition and field valida- Palmer, A.N., Palmer, M.V., and Sasowsky, I.D. (eds.), 1999, tion: 3rd Conference on hydrology, ecology, monitoring and Karst Modeling: Special Publication 5, Karst Waters Institute, management of ground water in karst terranes, Nashville, Akron Ohio. USA.

Robinson, P.C., 1984, Connectivity, flow and transport in net- Teutsch, G. and Sauter, M., 1998, Distributed parameter mod- work models of fractured media: Ph.D. Thesis, St. Catherine’s eling approaches in karst-hydrological investigations: Bulletin College, Oxford University du Centre d’Hydrogéologie, Neuchâtel, v. 16, p. 99-109.

Sawada, A, Uchida, M., Shimo, M., Yamamoto, H., Takahara, Torrence, C. and Compo, G., 1998, A practical guide to wavelet H., and Doe, T. W., 2000, Non-sorbing tracer migration experi- analysis: Bulletin of the American Meteorological Society, v. ments in fractured rock at the Kamaishi Mine, Northeast Japan: 79, p. 61-78. Engineering Geology, v. 56, p. 75-96 Wang, H.F. and Anderson, M.P., 1982, Introduction to ground- Shapiro, A. M. and Hsieh, P., 1994, Overview of research at the water modeling. W.H. Freeman and Co., San Francisco, USA Mirror Lake site: use of hydrologic, geophysical, and geochemical methods to characterize flow and transport in fractured Witherspoon, P.A., Wang, J.S.Y., Iwai, K., and Gale, J.E., 1980, rock: In U.S. Geological Survey Toxic Substances Hydrology Validity of the cubic law for fluid flow in a deformable rock Program – Proceedings of the Technical Meeting, Colorado fracture: Water Resources Research, v. 16, p. 1016-1024 Springs, Colorado, September 20-24, 1993, Morganwalp, D.W. and Aronson, D.A., (eds.), U.S. Geological Survey Water Re- sources Investigation Report 94-4015.

26 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Geochemistry and Climate Change

Geochemistry and Climate Change

Jay L. Banner1, MaryLynn Musgrove2, Jessica Rasmussen1, Jud Partin3, Andrew Long4, Brian Katz5, Barbara Mahler2, Larry Edwards6, Kim Cobb3, Eric James1, Russell S. Harmon7, Ellen Herman8, Carol M. Wicks9

1. Jackson School of Geosciences, University of Texas, Austin, TX 78712; 2. U.S. Geological Survey, Water Resources Division, Austin, TX 78754; 3. School of Earth and Atmospheric Sciences, Georgia Institute of Technology, Atlanta, GA 30332; 4. USGS South Dakota, Rapid City, SD 57702; 5. USGS Florida, Tallahassee, FL 32310; 6. Dept Geology and Geophysics, University of Minnesota, Minneapolis, MN 55455; 7. Department of the Army, Research Triangle, NC 27709; 8. Dept. Geology, Bucknell Uni- versity, Lewisburg, PA 17837; 9. Department of Geological Sciences, University of Missouri, Columbia, MO 65211.

INTRODUCTION Challenges

arst terrains are developed on soluble rock and typi- Karst aquifers pose many challenges to understanding the com- cally contain vital water resources. Over a range of plexities of their unique triple-porosity structure and tracing Ktime scales, chemical and clastic sedimentary deposits flow from recharge through infiltration to the different diffuse, accumulate in karst terrains. These deposits can provide re- fracture, and conduit flow paths within individual aquifers. In cords of past environments, including climate and hydrologic addition, the critical vulnerability of karst water resources to change, which have been shown to be of particular importance contamination due to their typically thin soil cover and discrete to other sciences and to social policy and planning. This paper recharge features underscore the importance of addressing reviews the state-of-the-art and prospects for future research on these challenges. Geochemical tracing of fluid flow paths in three different but related facets of the geochemistry of karst these systems is needed for testing and refining physical flow terrains: (i) The application of geochemistry to understanding models. Among the major issues faced by karst researchers the hydrogeology of karst aquifers, and (ii) the geochemistry and water resource managers are (i) determining the sources of of speleothems and other karst sedimentary deposits as records contaminants to groundwater and surface water; (ii) constrain- of past environmental change over different temporal scales ing the sites, timing, and amount of recharge to karst aquifers; that are preserved in these deposits, and (iii) climate change (iii) defining and quantifying groundwater-surface water inter- questions that may be addressed by speleothem studies. actions, and (iv) understanding the timing and thresholds of urbanization impacts. HYDROGEOCHEMISTRY Prospects Naturally-occurring and introduced chemical tracers have been applied over the past several decades to advance our under- Among the prospects for meeting the challenges outlined above standing of the hydrologic functioning of karst aquifers. In the are a number of naturally-occuring geochemical tracers, as dis- future, karst water resources face increased impacts, both in cussed below. terms of water supply demand and water quality, due to climate change and increases in population, urbanization, soil loss and Tracing flow paths: Identifying conduit flow paths in karst other land use changes. It will therefore become increasingly flow systems has been traditionally approached by different dye important to supplement conventional tracing approaches that tracing approaches. Deconvoluting the interaction between the use dyes, temperature, specific conductivity, ion concentrations conduit and diffuse end-members in a karst flow system can be and stable isotopes of H, O and C, via innovative ways to apply aided by geophysical (Vouillamoz et al., 2003), biological (e.g., these conventional geochemical tracers and by developing new Rossi et al., 1998; Mahler et al., 2000), and geochemical trac- tracers based on rare-earth elements, microbes and bacterio- ing approaches. Geochemical approaches include the analysis phages, stable isotopes of N S, and Fe, radiogenic isotopes of of (i) isotopic and chemical tracers that have distinct composi- Sr, Nd, U, Th and Pb, naturally-occurring particulate materials, tions in surface runoff, soils and aquifer bedrock, as this pro- chlorinated and fluorinated compounds, and hormones, phar- vides a measure of the extent of water-rock interaction within maceutical and personal care products. different reservoirs, and therefore a measure of the extent of interaction with soils and rocks, and of residence times in the

Frontiers of Karst Research 27

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Geochemistry and Climate Change

soil zone, in the conduit network, and in the matrix (Banner et groundwater over time can reveal aquifer dynamics such as the al., 1996; Uliana and Sharp 2001; Redwine and Howell, 2002; interaction of conduit and diffuse flow components. Celle-Jeanton et al., 2003; Wilcox et al., 2004); (ii) isotopic and chemical tracers with distinct compositions during periods of high vs. low recharge (Lakey and Krothe, 1996; Lee and Krothe, 2001; Jones and Banner, 2003), and (iii) the sediment dynamics in a karst system (Bottrell et al., 1999). Time series of geochemical tracer data are useful in karst for examining signals and responses of hydrologic events using methods such as breakthrough curve decomposition, frequency distribution analysis, and lumped-parameter models. (e.g., Dreiss, 1983; Larocque et al., 1999; Labat et al., 1999; Bouchaou et al., 2002). Such approaches can be complemented by using geochemical tracer methods in conjunction with lumped-parameter, physics- based, and/or mathematical (e.g. stochastic-based) hydrological modeling (Dreiss, 1983; 1989; Bonacci 1993; Eisenlohr et al., 1997; Halihan and Wicks, 1998; Jeannin, 2001; Long and Putnam, 2004). These methods are useful for understanding Figure 2. Conductance frequency distribution for Barton the complex dynamics of karst systems and quantifying the dif- Springs for 2000 (bold line) and the underlying normally dis- ferent flow components (Figs. 1, 2). There is significant poten- tributed populations (light lines). Time series of specific con- tial for innovative approaches using these methods in karst. ductance (electrical conductivity) are shown as inset. From Massei et al. (2007).

Figure 1. Transfer functions for well HH in the Madison Aquifer, South Dakota, representing combined conduit and intermediate components. All four transfer functions use a lognormal Figure 3. Probability density function of groundwater age from distribution for the conduit component and one of four distri- well RC-11 in the Madison Aquifer, South Dakota, using av- butions for the intermediate component, which includes log- erage parameter values calibrated for 1993-2005 (from A.J. normal, gamma, Pearson type III, or linear slope. The vertical Long and L.D. Putnam, work in progress). axis represents the fraction of flow for each time step in days for the combined conduit and intermediate flow. From Long and Putnam (2004). Determining contaminant sources and using contaminants as tracers: Building on an improved knowledge from tracing Groundwater age constraints: The distribution of ground- groundwater flowpaths, a multi-parameter approach will be- water age in a sample can be revealing regarding an aquifer’s come invaluable for determining the sources of contaminants to properties. The combined use of (i) multiple methods that con- karst flow systems. This will involve the combined use of dye- strain groundwater ages, such as 3H, 14C and CFC’s, and (ii) tracing studies, natural- and labeled-sediment tracers (Mahler new approaches to model the age information, has much poten- et al., 1998), environmental tracers (e.g., nutrients, isotopes, tial (Long and Putnam, 2004; in prep.; Fig. 3). Such approach- CFC’s, and bacteria; e.g., Mahler et al., 2000), pharmaceutical es permit examination of how changes in the age distribution of compounds and personal care products, and geophysical data

28 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Geochemistry and Climate Change

to identify contaminant sources and pathways of contaminant estimation of the uncertainty in these parameter estimates dif- transport in ground water, spring water, and surface water. Add- ficult, if not impossible. If such models can be advanced to ing to this arsenal, as analytical methods are advanced, will be overcome these uncertainties, then application of these tech- microbiological indicators as new tracers to track contaminant niques may be useful for more efficient objective quantification sources and movement in ground water from waste disposal of karst aquifer properties and data processing, as well as for and water reuse. advanced decision support systems in karst aquifers (Pierce et al., 2005). In contrast to determining sources of particular contaminants, some contaminants already in the environment may be used Using multiple geochemical indicators to estimate the timing effectively as tracers. Contaminants such as herbicides, insec- and amount of mixing of water from the epikarst with ground ticides and organic solvents offer significant advantages over water in shallow and deep parts of karst aquifers will be par- introduced artificial tracers. They represent the actual contami- ticularly important for public water supply wells. Tracers that nants of interest, typically are inadvertently introduced into the vary in composition seasonally, such as H and O isotopes, can system at the same time at multiple points, and they may be be used to determine the amount and extent of seasonality of introduced into the karst aquifer from a soil zone source dur- recharge (Jones et al., 2000; Andreo et al., 2004). Develop- ing storms, when contaminant transport is of the most interest ment of more extensive sampling/monitoring networks that (Fig. 4). quantify surface water, recharge and subsurface flow will allow such methods to be applied on an aquifer- or watershed-specific basis. Innovative application and statistical analysis of tracers as conventional as conductivity have the potential to delineate storm-flow from pre-storm water and to separate water types into different modes of aquifer functioning and response to climatic variations (Fig. 1).

Impacts of urbanization: Geochemical tracers have great potential for quantifying the sources of urban water to karst and the impacts of urbanization on karst systems. Urbanization has been variably proposed to divert or concentrate recharge, via increased runoff, and, in contrast, to increase recharge, through Figure 4. Decomposition of atrazine breakthrough from Bar- leaking urban infrastructure (Krothe et al., 2002; Pierce et al., ton springs, Austin, Texas, in response to a storm. Atrazine 2004). The impacts of land use change and urbanization on breakthrough (data shown as open circles) is modeled as the sum (bold line) of three log-normal distributions (fine lines) rep- contaminant vulnerability (nutrients such as nitrate, pesticides, resenting three different inputs of atrazine. Peaks in the indi- volatile organic compounds, pharmaceuticals and personal care vidual breakthrough curves align well with troughs in specific products) and the non-linearity of processes within karst sys- conductance (grey line). From Mahler and Massei (2006). tems require that the thresholds of these impacts be understood. Geochemical time series that delineate the onset and extent of Groundwater management: It would be of great benefit to these impacts relative to the extent of urban development can be groundwater managers and environmental consultants to have used to (i) identify impact thresholds and (ii) delineate between improved information about karst systems, such as (i) the tim- the contrasting effects on recharge. Such time-series data may ing, seasonality, and spatial distribution of recharge, (ii) the role come from recent karst spring deposits such as travertine (e.g., of epikarst in storage and recharge to the karst aquifer (Bonac- DeMott et al., 2006). ci, 2001), (iii) the impact of brush clearing and deforestation on recharge, (iv) estimating aquifer yield (Bacchus, et al., 2003), Research Infrastructure Needs (v) groundwater-surface water interactions, and (vi) aquifer response to anthropogenic impacts and climate change. Within Analytical advances in terms of isotopic analysis of specific or- the hydrologic community, advanced computer modeling and ganic compounds in water, portable chemical methods such as numerical simulation programs are available to efficiently and laser-induced breakdown spectroscopy (Harmon et al., 2005), objectively estimate model parameters. These tools are dif- Raman spectroscopy and mini-differential optical absorption ficult to apply, however, to karst systems due to the extreme spectrometry can be applied to the in-situ monitoring of karst heterogeneities in recharge and the subsurface, which make the flow systems and the deposits that form therein. To the extent

Frontiers of Karst Research 29

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Geochemistry and Climate Change

that the controls of environmental and anthropogenic factors ter that occurs during calcite precipitation. Water dripping from

on geochemical variations of modern karst systems can be un- the cave ceiling degasses CO2 and deposits CaCO3 as stalactites derstood, the more accurate will be interpretations of the geo- and then stalagmites and flowstone as the water subsequently chemical records preserved in karst deposits (e.g., Tooth and reaches the cave walls and floor. Stalagmites grow at rates in Fairchild, 2003), as discussed in the next sections. the range of tens to hundreds of μm/year, so that a 1-m long stalagmite may represent a deposition time of 104-105 years. Age SPELEOTHEMS AS profiles of stalagmites can best be determined by U-series geo- PALEOENVIRONMENTAL RECORDS chronologic techniques, although other dating techniques have been applied to speleothems. Axial profiles of stable isotope Caves are geographically widespread features within karst composition (e.g., calcite 18O/16O and 13C/12C and fluid inclu- landscapes and provide unique preservation environments sion D/H and 18O/16O), trace element abundances, calcite color for deposits that contain proxy records of past environmental and luminescence banding, and organic acid content can pro- change. Speleothems – the chemical precipitates that form in vide robust information about the environment of present and caves – are especially important repositories for many different past speleothem deposition as well as surface climate on local, types of geochemical information. Research in the 1970s and regional, and global scales. 1980s demonstrated clearly that speleothems can be dated and contain information, which under certain circumstances, can be Challenges used to infer the conditions of their environment of deposition (Hendy, 1971; Thompson et al., 1975; Schwarcz et al., 1976; There are several research needs and challenges related to un- Harmon et al., 1978; Gascoyne et al., 1983; Yonge et al., 1985). derstanding the formation of speleothems in caves and the use More recent research undertaken during the past two decades of the geochemical information contained in speleothems. The has illustrated that variations in the isotopic and chemical sig- presence of humic, fulvic, and other organic acids has been nal contained in speleothems can be used to reconstruct past documented in speleothems and are the main source of calcite changes in karst terrain hydrologic behavior as well as surface color and luminescence. Depositional controls on organic com- vegetation, weather, and climate change over different spatial pounds in speleothems are not well known and the paleoclimate and temporal scales. significance of variations in organic compound abundance and composition in speleothems remains to be established. Varia-

The primary speleothem mineral is calcite, CaCO3, but its tions in speleothems of trace elements such as Mg, Sr, Zn, Mn, metastable polymorph aragonite also occurs in the cave en- Pb and U are well documented, but their potential as monitors vironment. In addition, other carbonate and sulfate minerals, of environmental change are yet to be fully determined. An-

such as gypsum, CaSO4·2H2O, can be found in some cave en- other important need is an improved understanding that will vironments. Less common minerals are found in special cave lead to improving the models, or ‘transfer functions’, through environments, usually in small amounts (Hill and Forti, 1997). which the parameters measured in speleothem calcite are trans- Carbonate and sulfate minerals take on a variety of forms, but lated into first-order climate information (e.g. variations in stalagmites, stalactites and flowstones are the most useful envi- speleothem O-isotope ratios to surface temperature changes or ronmental archives. C-isotope composition to surface vegetation type). In addition to speleothems, clastic sediment deposits also have potential to Rainfall, with an H- and O-isotope signature determined by its be dated (Schmidt, 1982; Granger and Muzikar, 2001; Granger origin and transport history in the atmosphere, falls on the karst et al., 2001; Anthony and Granger, 2004) and provide useful land surface and infiltrates into the epikarst zone. With a vari- paleoclimate information (Zhang, 1998). able time delay that is dependent on the complexity of the subsurface flow path, the water then percolates through the vadose Prospects zone until it intercepts an underlying cave. Chemical equili- bration and ion exchange in the soil, chemical reactivity at the Caves are widespread features in karst terrains and speleothems soil-bedrock contact, and the subsequent physical and chemi- are common cave features. Therefore, once fully understood, cal reactions occurring during infiltration through the limestone speleothem geochemical records have the potential to become and during calcite precipitation once the seepage water enters the most important paleoenvironmental archive for carbonate the cave ultimately determines the trace chemical and isotopic islands and terrestrial domains, particularly for continental in- composition of a speleothem. Of particular importance is the teriors. stable isotope exchange between cave atmosphere and drip wa-

30 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Geochemistry and Climate Change

Research Infrastructure Needs this has been a continuing and underappreciated challenge for speleothem studies (Mickler et al., 2004, 2006). In short, the There also are specific infrastructure needs that would facilitate challenges described here may be summarized by the question: the use of speleothems as paleoenvironmental records. Most Can speleothems serve as equivalents of ice cores and deep-sea critical is a formal system for access to state-of-the-art U-series sediment cores as records of environmental change for tropical, geochronology because, at present, the number of age deter- temperate, and ice-free high-latitude terrestrial domains? minations needed by the research community is stressing the informal collaborative network that has been providing such support over the past decade. A central database of speleothem paleoenvironmental data would help to elucidate climate patterns at the regional or continental scale at different times. An archive for speleothems used for paleoenvironmental study would support the conservation of these rare materials, facilitate community access to studied samples for other types of analysis, and preserve these rare materials for future use as new analytical techniques emerge.

CLIMATE CHANGE

Speleothems can provide records of past cave depositional environment, karst aquifer hydrologic flow regime, and surface 230 vegetation character, weather, and climate. Among the prox- Figure 5. Comparison of analytical precision, Th age, and sample size (in quantities of 238U) for different U-series ana- ies in speleothems that are used in this regard are growth tim- lytical techniques; α-spectrometry, thermal ionization mass ing and rate, calcite stable isotopes (C and O) and radiogenic spectrometry, and inductively-coupled mass spectrometry. isotope (Sr and U) compositions, trace element variations, and The main reason for the improvement is the increase in ionization/transmission efficiency by about an order of magnitude fluid inclusion isotopic (H and O) composition. (to more than 1%). The lines are calculated on the basis of counting statistics, but it has been demonstrated that the cal- Challenges culated precisions are possible on real samples. (From the work of R.L. Edwards, 2007). Understanding the temporal variability of water flow and storage in karst systems is needed over a range of time scales, including (i) seasonal to decadal scales, which are important for Prospects managing water resources, water quality and water availability, and (ii) century to millennial time scales, which are important Geochronology: Any record of past environmental change, be for reconstructing hydrologic and climate change and their in- it contained in a karst deposit, a marine sediment record or in an teractions. One challenge is constructing records for key time ice core, is only as good as its temporal resolution. It follows periods including the last glacial and interglacial maxima, in- that accurate and precise geochronologic methods are the key tervals of abrupt change such as the Pleistocene-Holocene tran- to developing useful speleothem geochemical proxy records. sition, the Younger Dryas, and the Little Ice Age. In addition, In addition, non-geochemical variations such as the timing and records are needed that cover the last century, to coincide with growth rate of speleothem calcite can serve as an additional instrumented climate records and as a basis of predicting near- proxy if high-resolution chronologic data can be obtained. The term future trends. An appropriate spatial coverage of climate advancing state-of-the-art is promising in this regard. Spele- records is needed in key geographic areas that drive and those othem records can be continuous, spanning long periods of that respond to important climate phenomena, such as the El time, up to hundreds of thousands of years, and can be analyzed Niño-Southern Oscillation and the seasonal monsoons. As with at high temporal resolution, often as high as decades and some- all proxy records of past change, valid interpretations of spele- times higher. Major improvements in analytical measurements othem records are required. For speleothems, it is necessary to of U-series nuclides took place two decades ago (Edwards et determine if equilibrium isotope and trace element precipita- al., 2003; Richards and Dorale, 2003), and additional improve- tion occurs during the precipitation of individual calcite lay- ments continue to occur (Musgrove et al., 2001; Goldstein and ers. Although recognized by early workers (e.g. Hendy, 1971), Stirling, 2003; Edwards, 2007). Low porosity speleothem cal-

Frontiers of Karst Research 31

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Geochemistry and Climate Change

cite that is free of detritus can be dated precisely and accurately annual transitions and seasonality in the middle-late Holocene by U/Th/Pa dating methods using modern mass spectrometry including modern and historical periods such as the Little Ice (Fig. 5). For less pure samples, geochronologic analysis of Age. The construction of records that cover the past century multiple speleothems from the same cave system can be used will increase the applicability of speleothem records to the to develop convergent age models (Fig. 6). Some speleothems current issue of global warming. Among the climate change also exhibit repetitive banding, which can provide highly pre- mechanisms of interest are: long-term astronomical forcing cise relative ages (Tan et al., 2006). and different recurrence intervals; variability in solar intensity; atmospheric-ocean oscillations (e.g., El Niño-Southern Oscil- lation, Pacific Decadal Oscillation, North Atlantic Oscillation); variability in temperature, precipitation, and monsoon intensity; and anthropogenic climate forcing.

Individual weather events: Among other measures of climate, the O-isotopic composition of cave calcite can provide a measure of fluctuations in the 18O/16O ratio of meteoric precipitation, an important variable that reflects the global climate state. For example, during tropical cyclone formation, intense rainfall will cause atmospheric water vapor masses to evolve to very low δ18O values. An intriguing 20th century stalagmite record from Belize (Frappier et al., 2007), sampled at very high spatial resolution (20 micron steps) reveals negative oxygen isotope excursions that coincide with the historical record of hurricanes in the Belize region (Fig. 7).

Figure 6. Three stalagmite records from Borneo. There were a total of five rounds of drilling samples and dating. After every round of dating, the stalagmite records continued to converge on a "master-record" which resembled the others. For this study, this required eliminating portions of records growing <10µm/yr (light portions on figure) (Partinet al., 2007)

Key time-scales and regions: As noted above, paleo-records of seasonal-decadal and century-millennial resolution are needed. The research community should define several key geographic regions to be studied for a given climate mechanism and its impacts. For example, El Niño-Southern Oscillation phenomena are driven by sea surface temperatures in the Western Pacific Warm Pool, whereas its impacts occur in many regions around the globe, including the arid coast of Andean South America and Australia’s drought-prone watersheds. Through the use of Figure 7. A Belizean stalagmite stable isotope record (B) and multiple proxies, such as speleothems, pollen, mollusks, corals, its correlation with recent tropical cyclone events in the re- and tree rings) in a specific region and over a particular time gion of Belize (A), and the Southern Oscillation Index (C). From Frappier et al., (2007). interval, the different proxy records can be used to assess each other’s efficacy and applicability. Abrupt climate change: Marine deep-sea sediment core and Key time periods: In order to address some fundamental ques- ice core paleoclimate studies have demonstrated that late Qua- tions about climate change mechanisms, it is necessary to in- ternary climate was dominated by abrupt climate change, geo- crease speleothem sampling resolution within and across the graphically large and widespread shifts in climate taking place following climatic periods: glacial-interglacial transitions such over just a decade (NRC, 2004). A significant societal concern as the Last Glacial Maximum and the Younger Dryas; inter- stemming from this work is the possibility that greenhouse gas

32 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Geochemistry and Climate Change

driven warming could trigger such an abrupt shift in the coming fers and the modern aquifer system are going to be essential decades or centuries. Cave climate studies have the potential to for interpreting temporal records. This will require studies of provide valuable information on the spatial extent and causes changes 1) above caves (thickness, moisture, water chemistry, of abrupt climate shifts in the past (Fig. 8). productivity of soils, vegetation, and weather, 2) at cave drip sites (cave-air meteorology, drip rate, drip composition, and Climate change summary: Several major contributions are 3) event sampling. Studies incorporating changes in land use expected from speleothem climate research. (i) A precise and driven by humans will make speleothem studies more relevant accurate chronology for climate change for the last 600,000 to human time scales, all independent of speleothem studies so years will be established. In the next several years, it is likely that speleothem records may be better understood. This will that climate records with chronologies precise to within a cen- require an even more extensive monitoring system and research tury will be obtained for events within the last 100,000 years. infrastructure funding than that described above in the Geo- These chronologies will ultimately be based on concordant U- chemistry section. series dates for speleothems and correlations between speleothems and other types of proxy climate records. (ii) This line of research will yield a high-resolution picture of the temporal and spatial patterns of climate change and (in conjunction with ice core records) changes in atmospheric gas composition over the past several hundred thousand years. (iii) Based on these records, more comprehensive ideas will be formulated about the causes of climate change in the past, and by extension, into the future. The initial work in this field has demonstrated a strong connection between atmospheric composition and the glacial-interglacial cycles of the late Quaternary (Fig. 8).

Figure 9. Seasonal variations in speleothem calcite growth on

artificial substrates (red) and cave-air CO2 (grey) in a central Texas cave. From Banner et al., (2007).

The road to high-fidelity, well-dated speleothem paleoclimate records is paved with more dates than at first appearance are necessary (Fig. 6), and more reproducibility than is often funded/published. This will require laborious ultra-high-resolution reconstructions of 20th century climate and, absent visible banding, identifying chemical, isotopic, or petrographic proxies that occur in annual cycles, and counting back these annual Figure 8. Correlation of Asian Monsoon, as recorded in a sta- cycles. The methodology for executing such records exists or lagmite from Hulu Cave in China, to ice core methane and can be developed, but it will require significant effort that can oxygen isotopes, and to marine core Heinrich Events H1 through H4 (Wang et al., 2001). be driven by funding priorities.

To improve the accuracy of U-series age determinations, we Research Infrastructure Needs will need an improved understanding of initial 230Th/232Th values incorporated into speleothems. Studies of modern system, Evaluation and calibration of speleothem proxies, including the zero-age speleothems, akin to studies of living corals to con- assessment of equilibrium precipitation of speleothem calcite, strain marine initial 230Th/232Th, will be valuable for addressing will be achieved through two main methods: (1) analysis of this issue. multiple proxies over a common time interval in the past; and (2) analysis of the modern system via monitoring networks in In the next decade, this field of study will have a major impact karst aquifers (Mickler et al., 2004, Fig. 9). Understanding of in two areas of the earth sciences: (1) the understanding of the temporal changes of the modern landscape above karst aqui- major factors that have caused the earth’s climate to change in

Frontiers of Karst Research 33

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Geochemistry and Climate Change

the past, and that likely will cause the earth’s climate to change cosmogenic 26Al and 10Be: Journal of Cave and Karst Studies, in the future, and (2) the understanding of those soil, vadose v. 66, p. 46-55. zone, cave, and climatic processes that affect the isotopic and Bacchus, S.T., Archibald, D.D., Brook, G.A. Britton, K.O., chemical composition of karst waters and the cave deposits that Haines, B.L., Rathbun, S.L. and Madden, M., 2003, Near-in- precipitate from them. Important climate science goals will frared spectroscopy of a hydroecological indicator: new tool relate to the first point, whereas some significant understand- for determining sustainable yield for Floridan aquifer system: ing of the second set of processes will be necessary in order to Hydrological Processes, v. 17, 1785-1809. fully realize the first set of goals. The improved understanding of the second set of processes will have direct application to Banner, J.L., Musgrove, M., Asmerom, Y., Edwards, R.L. and important science questions regarding the operation of modern Hoff, J.S., 1996, High-resolution temporal record of Holocene karst systems. ground-water chemistry: Tracing links between climate and hydrology: Geology, v. 24, p. 1049-1053.

RELATION TO OTHER FIELDS OF STUDY Banner, J.L., Guilfoyle, A., James, E.W., Stern, L.A., and Mus- grove, M.L, 2007, Seasonal variations in modern speleothem Other fields of study that karst research would have synergy calcite growth in central Texas, U.S.A.: Journal of Sedimentary with and bearing on are landscape ecology, microbial ecology; Research, v. 77, p. 615-622. the global carbon cycle; biogeochemical cycles; water resource policy, contaminant flow and transport, and decision support Bonacci, O., 1993, Karst springs hydrographs as indicators of karst aquifers: Hydrological Sciences Journal, v. 38, p. 51-62. systems. Bonacci, O., 2001, Analysis of the maximum discharge of karst IMPLICATIONS: CAVE CONSERVATION springs: Hydrogeology Journal, v. 9, p. 328-338.

The challenges are: Can we work in these environments with- Bottrell, S., Hardwick, P. and Gunn, J., 1999, Sediment dynam- out adversely affected them? Is it possible to enact an inter- ics in the Castleton karst, Derbyshire, UK: Earth Surface Pro- national research policy regarding cave research that scientists cesses and Landforms, v. 24, p. 745-759. will follow? The varying factors of different local and federal Bouchaou, L., Mangin, A. and Chauve, P., 2002, Turbidity restrictions on cave access and research permitting is an ob- mechanism of water from a karstic spring: example of the Ain stacle to be addressed. One aspect of such a policy might be Asserdoune spring (Beni Mellal Atlas, Morocco): Journal of the coring and reconnaissance analysis of speleothems vs. the Hydrology, v. 265, p. 34-42. multiple whole sample removal from caves. The karst community lacks an archival system for researchers such as that Celle-Jeanton, H., Emblanch, C., Mudry, J. and Charmoille, 2+ which exists for ice cores, yet speleothems are rarer than any A., 2003, Contribution of time tracers (Mg , TOC, delta C-13(TDIC), NO - ) to understand the role of the unsatu- individual ice core. There are competing pressures of multiple 3 rated zone: A case study - Karst aquifers in the Doubs val- sampling trips needed if reconnaissance sampling is employed, ley, eastern France: Geophysical Research Letters, v. 30, doi vs. the time and funding required for the reconnaissance ap- 1029/2002GL016781. proach. One way to enact such a policy, if it could be agreed upon, would be to require a conservation plan as part of the Cooke, M.J., Stern, L.A., Banner, J.L., Mack, L.E., Stafford, T., “Broader Impacts” section of a research grant proposal. and Toomey, R.S., 2003, Precise timing and rate of massive late Quaternary soil denudation: Geology, v. 31, p. 853-856. REFERENCES DeMott, L. M., Banner, J., and Christian, L., 2006, Recent travertine deposits as records of groundwater processes in urban- Andreo, B., Liñán, C., Carrasco, F., Jiménez de Cisneros, C., izing environments: Geological Society of America Abstracts Caballero, F. and Mudry, J., 2004, Influence of rainfall quan- with Programs, v. 38 p. 289. tity on the isotopic composition (O-18 and H-2) of water in mountainous areas. Application for groundwater research in the Dreiss, S. J., 1983, Linear unit-response functions as indicators Yunquera-Nieves karst aquifers (S Spain): Applied Geochemis- of recharge areas for large karst springs: Journal of Hydrology, try, v. 19, p. 561-574. v. 61, p. 31-44.

Anthony, D.M. and Granger, D.E., 2004, A Late Tertiary ori- Edwards, R.L., Gallup, C.D. and Cheng, H, 2003, Uranium- gin for multilevel caves along the western escarpment of the series dating of marine and lacustrine carbonates: Reviews in Cumberland Plateau, Tennessee and Kentucky, established by Mineralogy and Geochemisry, v. 52, p. 363-405.

34 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Geochemistry and Climate Change

Eisenlohr, L., Király, L., Bouzelboudjen, M. and Rossier, Y., Jones, I. C. and Banner, J. L., 2003, Hydrogeologic and climat- 1997, Numerical simulation as a tool for checking the inter- ic influences on spatial and interannual variation of recharge to pretation of karst spring hydrographs: Journal of Hydrology, v. a tropical karst island aquifer: Water Resources Research v. 39, 193, p. 306-315. p. 1253-1263.

Frappier, A.B., Sahagian, D., Carpenter, S.J. González, L.A. Jones, I.C., Banner, J.L. and Humphrey, J.D., 2000, Estimating and Frappier, B., 2007, A stalagmite proxy record of recent recharge in a tropical karst aquifer: Water Resources Research, tropical cyclone events: Geology, v.7, p. 111–114. v. 36, p. 1289-1299.

Gascoyne, M., Schwarcz, H.P. and Ford, D.C., 1983, Uranium- Krothe J.N., Garcia-Fresca B., Sharp J.M. Jr, 2002, Effects of series ages of speleothem from Northwest England: Correla- urbanisation on groundwater systems: In Proceedings of the tion with Quaternary climate: Philosophical Transactions of the XXXII IAH and VI ALHSUD Congress on Groundwater and Royal Society of London, v. B301, p. 143-164. Human Development, Mar del Plata, Argentina.

Goldstein, S.J. and Stirling, C.H., 2003, Techniques for mea- Labat, D., Ababou, R. and Mangin, A., 1999, Linear and non- suring uranium-series nuclides: 1992-2002: Reviews in Miner- linear input/output models for karstic springflow and flood alogy and Geochemistry, v. 52, p. 23-57. prediction at different time scales: Stochastic Environmental Research and Risk Assessment, v.13, p. 337-364. Granger, D.E. and Muzikar, P., 2001, Dating sediment burial with in situ-produced cosmogenic nuclides: theory, techniques, Lakey, B.L. and Krothe, N.C., 1996, Stable isotopic variation and limitations: Earth and Planetary Science Letters, v. 188, p. of storm discharge from a perennial karst spring, Indiana: Wa- 269-281. ter Resources Research, v. 32, p. 721–731.

Granger, D.E., Fabel, D., and Palmer, A.N., 2001, Pliocene- Larocque, M., Banton, O., Ackerer, P. and Razack, M., 1999, Pleistocene incision of the Green River, Kentucky, determined Determining karst transmissivities with inverse modeling and from radioactive decay of cosmogenic 26Al and 10Be in Mam- an equivalent porous media: Ground Water, v. 37, p. 897-903. moth Cave sediments: GSA Bulletin, v. 113, p. 825-836. Lee, E.S. and Krothe, N.C., 2001, A four-component mixing Halihan, T. and Wicks, C.M., 1998, Modeling of storm re- model for water in a karst terrain in south-central Indiana, sponses in conduit flow aquifers with reservoirs: Journal of USA using solute concentration and stable isotopes as tracers: Hydrology, v. 208, p. 82-91. Chemical Geology, v. 179, p. 129-143.

Harmon, R.S., DeLucia, F.C., McManus, C.E., McMillan, N.J., Long, A.J. and Putnam, L.D., 2004, Linear model describing Jenkins, T.F., Walsh, M.E., and Miziolek, A.W., 2005, Laser- three components of flow in karst aquifers using O-18 data: Induced Breakdown Spectroscopy - An Emerging Chemical Journal of Hydrology, v. 296, p. 254-270. Sensor Technology for Field-Portable, Real-Time Geochemi- cal, Mineralogical, and Environmental Applications: Applied Mahler, B.J., Bennett, P.C. and Zimmerman, M., 1998, Lan- Geochemistry, v. 21, p. 730-747. thanide-labeled clay: A new method for tracing sediment transport in karst: Ground Water, v.36, p. 835-843. Harmon, R.S., Schwarcz H.P., and Ford, D.C., 1978, Stable isotope geochemistry of speleothems and cave waters from the Mahler B. and Massei, N., 2006, Anthropogenic contaminants Flint Ridge-Mammoth Cave System, Kentucky: Journal of Ge- as tracers in an urbanizing karst aquifer, Journal of Contami- ology, v. 86, p. 373-384. nant Hydrology, v. 91, p. 81-106.

Hendy, C.H., 1971, The isotopic geochemistry of speleothems Mahler, B.J., Personne, J.C., Lods, G.F. and Drogue, C., 2000, –I. The calculation of the effects of different modes of forma- Transport of free and particulate-associated bacteria in karst: tion on the isotopic composition of speleothems and their ap- Journal of Hydrology, v. 238, p. 179-193. plicability as palaeoclimatic indicators: Geochimica et Cos- mochimica Acta, v. 35, p. 801-824. Massei N., Mahler B., Bakalowicz M., Fournier M., and Du- pont, J., 2007, Quantitative interpretation of specific conduc- Hill, C.A. and Forti, P., 1997, Cave Minerals of the World: 2nd tance frequency distributions in Karst: Ground Water, v. 45, p. Ed., National Speleological Society, Huntsville, AL, 463 p. 288-293.

Jeannin, P.Y., 2001, Modeling flow in phreatic and epiphreatic Mickler, P.J., Banner, J.L., Stern, L., Asmerom, Y., Edwards, karst conduits in the Hölloch Cave (Muotatal, Switzerland): R.L. and Ito, E., 2004, Stable isotope variations in modern trop- Water Resources Research, v. 37, p. 191-200. ical speleothems: Evaluating equilibrium vs. kinetic isotope effects: Geochimica et Cosmochimica Acta, v. 68, p. 4381-4393.

Frontiers of Karst Research 35

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Geochemistry and Climate Change

Mickler, P.J., Stern, L.A. and Banner, J.L., 2006, Large kinetic Tan, M., Baker, A., Genty, D., Smith, C., Esper, J. and Cai, isotope effects in modern speleothems: Geological Society of B., 2006, Applications of stalagmite laminae to paleoclimate America Bulletin, v. 117, doi 10.1130/B25698.1 reconstructions: comparison with dendrochronology/climatol- ogy: Quaternary Science Reviews, v. 25, p. 2103–2117. Musgrove, M.L., Banner, J.L., Mack, L.E., Combs, D.M., 234 238 James, E.W., Cheng, H. and Edwards, R.L., 2001, Geochronol- Thompson, P., Ford, D.C., and Schwarcz, H.P., 1975, U /U ogy of late Pleistocene to Holocene speleothems from central ratios in limestone cave seepage waters and speleothem from Texas: Implications for regional paleoclimate: Geological So- West Virginia: Geochimica et Cosmochimica Acta, v. 39, p. ciety of America Bulletin, v. 113, p. 1532-1543. 661-669.

National Research Council, 2004, Abrupt Climate Change: In- Tooth, A.F. and Fairchild, I.J., 2003, Soil and karst aquifer evitable Surprises: Committee on Abrupt Climate Change, Na- hydrological controls on the geochemical evolution of spe- tional Academy Press, Washington, D.C. leothem-forming drip waters, Crag Cave, southwest Ireland: Journal of Hydrology, v. 273, p. 51-68. Partin, J.W., Cobb, K.M, Adkins, J.F., Clark, B. and Fernandez, D.P., 2007, Millennial-scale trends in warm pool hydrology Uliana, M.M. and Sharp, J.M., 2001, Tracing regional flow since the last glacial maximum: Nature, v. 449, p. 452-455. paths to major springs in Trans-Pecos Texas using geochemical data and geochemical models: Chemical Geology, v. 179, Pierce, S.A., Sharp, J.M., Jr., and Garcia-Fresca, B., 2004, In- p. 53-72. creased groundwater recharge rates as a result of urbanization: Hydrological Science and Technology, v. 20, p. 119-127. Vouillamoz, J.M., Legchenko, A., Albouy, Y., Bakalowicz, M., Baltassat, J.M. and Al-Fares, W., 2003, Localization of satu- Pierce SA, Sharp J, Lowry T, Tidwell V., 2005, Decision sup- rated karst aquifer with magnetic resonance sounding and report theory and sustainable management of the karstic Edwards sistivity imagery: Ground Water, v. 41, p. 578-586. aquifer: Geological Society of America Abstracts with Pro- grams. v. 37, p. 7, 32. Wang,Y.J., Cheng, H., Edwards, R.L., An, Z.S., Wu, J.Y., Shen, C.-C. and Dorale, J.A., 2001, A High-Resolution Absolute-Dat- Redwine, J.C. and Howell, J.R., 2002, Geochemical methods ed Late Pleistocene Monsoon Record from Hulu Cave, China: for distinguishing surface water from groundwater in the Knox Science, v. 294, p. 2345-2348 Aquifer System: Environmental Geology, v. 42, p. 485-491. Wilcox, W.M., Solo-Gabriele, H.M. and Sternberg, L.O.R., Richards, D.A. and Dorale, J.A., 2003, Uranium-series chro- 2004, Use of stable isotopes to quantify flows between the Ev- nology and environmental applications of speleothems: Re- erglades and urban areas in Miami-Dade County Florida: Jour- views in Mineralogy and Geochemistry, v. 52, p.407-460. nal of Hydrology, v. 293, p. 1-19.

Rossi, P., Dorfliger, N., Kennedy, K., Muller, I. and Aragno, Yonge, C.J, Ford, D.C., Gray, J. and Schwarcz, H.P., 1985, Sta- M., 1998, Bacteriophages as surface and ground water tracers: ble isotope studies of cave seepage water: Chemical Geology, Hydrology and Earth System Sciences, v. 2, p. 101-110 v. 58, p. 97-105.

Schmidt, V.A., 1982, Magnetostratigraphy of Sediments in Zhang, D.D., 1998, A mineralogical analysis of karst sediments Mammoth Cave, Kentucky: Science, v. 217, p. 827-829. and its implications to the middle-late Pleistocene climatic changes on the Tibetan Plateau: Journal of the Geological Soci- Schwarcz, H.P., Harmon, R.S., Thompson, P., and Ford, D.C., ety of India, v. 52, p. 351-359. 1976, Stable isotope studies of fluid inclusions in speleothems and their paleoclimate significance: Geochimica Cosmochimi- ca Acta, v. 40, p. 657 665.

36 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Caves and Karst as Model Systems

Caves and Karst as Model Systems for Advancing the Microbial Sciences

Annette Summers Engel Department of Geology and Geophysics Louisiana State University E235 Howe-Russell Geoscience Complex Baton Rouge, LA 70803

Diana E. Northup Department of Biology University of New Mexico MSC03 2020 Albuquerque, NM 87131-0001

INTRODUCTION 2003). Microbes can also be flushed into caves from meteoric drip waters (Laiz et al., 1999), impacting the carbon load in nu- he subsurface is one of the Earth’s major general habi- trient depauperate ecosystems. However, because caves con- tats, defined as >8 m depth for terrestrial settings (e.g., tain a range of physiochemical environments and habitats, and Whitman, et al., 1998). With millions of estimated different physicochemical parameters dictate distinct metabolic T strategies by microorganisms adapted to life in that particular microbial species (i.e. prokaryotic) occurring worldwide (e.g., Dykhuizen, 1998; Curtis et al., 2002), the abundance of mi- set of parameters, there are also a wide variety of microbes that crobes in the subsurface could exceed numbers found else- have been identified from caves and karst recently using cul- where in the biosphere (Pace, 1997). However, our knowledge ture- and molecular-based techniques. of the types and extent of subterranean life is poor, as is our understanding of the biogeochemical roles of microbes in the One of the critical extreme conditions affecting most subsurface- subsurface, due in part to limited accessibility and difficulty adapted fauna, and the establishment and sustainability of sub- in obtaining uncontaminated material from the subsurface. surface, including cave, ecosystems is the lack of a continuous Caves, representing discontinuous habitats that are character- and rich nutrient supply, usually in the form of surface-derived ized by complete darkness, nearly constant air and water tem- (allochthonous) photosynthetically produced organic matter. peratures, and relative humidity near saturation, allow for easy In the absence of sunlight, however, redox-variable groundwa- access to the subsurface from which samples can be taken and ter and reactive rock surfaces provide a wide assortment of po- in situ experimentation is possible. tential energy sources that chemolithoautotrophic microbes can use to gain cellular energy. Some subsurface and cave ecosys- Caves can extend well below the 8 m depth, and link the sur- tems have been shown to rely exclusively on chemosynthesis face to the subsurface. Early speleologists showed that mi- (e.g., Stevens, 1997; Amend and Teske, 2005). The quality and crobes colonize caves, but their presence was predominately quantity of chemosynthetic energy result in karst ecosystems assumed to be as secondary degraders and food sources for teaming with life, from novel microbial groups (e.g., Engel et higher organisms (e.g., Caumartin, 1963). Microorganisms in al., 2003) to numerous endemic, cave-adapted metazoans (e.g., caves were considered to be merely a subset of surface, mostly Sarbu et al., 1996; Culver and Sket, 2000; Hose et al., 2000; soil, microbial communities, with just a few groups being po- Sarbu et al., 2000). The very recent discovery of the Ayalon tentially important for geological processes (e.g., Symk and Cave ecosystem in Israel, with eight invertebrates new to sci- Drzal, 1964; Hubbard et al., 1986). Indeed, recent research ence and the possibility of a new chemosynthetically-based has verified that cave and karst microbes are often translocated ecosystem, is testimony that substantial and unknown subter- soil heterotrophs, chemoorganotrophs, or fecal coliforms that ranean diversity still exists (Por, 2007) come into cave and karst systems from surface streams, on air currents, or carried into the cave by animal life, along with al- From a deep time perspective, the impact of microbes on cave lochthonous organic matter and nutrient-rich sediments (e.g., and karst habitats can be profound. Karst is ubiquitous world- Rusterholtz and Mallory, 1994; Mikell et al., 1996; Grothand wide (12.5% of the ice-free land surface), and carbonate rocks, Saiz-Jimenez, 1999; Khizhnyak et al., 2003; Simon et al., accounting for ~20% of all Phanerozoic sedimentary rocks,

Frontiers of Karst Research 37

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Caves and Karst as Model Systems

form today and exist in 3.5 Ga deposits (e.g., White, 1988; Ford tigations, most research is at the descriptive stage. Only a few and Williams, 2007). Therefore, karst surface and subsurface species have been described from cultures, and only recently microbial habitats have presumably existed throughout much have non-invasive genomics methods been employed. By con- of Earth’s history. Moreover, because of physical constraints, necting phylogeny to function utilizing cultivation, genomics, caves and karst can be long-term reservoirs for microbial com- and proteomics in the future, we will have an improved view of munities, as some systems remain relatively unchanged for the ways that microbes shape their habitat and of the role that thousands, if not millions, of years (e.g., Gale, 1992). the cave and karst habitat has in selecting the metabolic and functional types of microbes within that habitat. Considering the worldwide distribution of karst, and regardless of the public’s fascination with “dark life” and unseen micro- In this overview, we summarize the current knowledge base bial worlds (e.g., Taylor, 1999), or of appreciating the sheer regarding microbial sciences of cave and karst settings, present biodiversity of life on earth, it is imperative that we obtain a basic research questions surrounding the study of microbes in better understanding of the types of microbial communities in caves and karst, and explore the relevance of these investiga- these systems, as well as what physicochemical and ecologi- tions to advancing disciplinary fields within the microbial sci- cal conditions may be controlling diversity. These basic geo- ences. We attempt here to recognize international efforts, but biological issues are critical to identifying how microbes alter acknowledge that most of the review emphasizes North Ameri- karst and affect the productivity of karst aquifers and petro- can and European work published in English. The foci of the leum reservoirs (e.g., Hill, 1990; Palmer, 1995; Mahler et al., review, in the context of caves and karst, are: 2000), and to preserving the integrity of karst through predicting changes that may occur following disturbance (van Beynen • Microbial biodiversity and Townsend, 2005). • Microbes and ecosystem function • Microbial roles in geochemical and geological processes In general, new molecular techniques developed in the 1980s and ‘90s increased the number of environments that could be We conclude with possible applied aspects of cave and karst successfully studied by microbiologists (Pace, 1997). These geomicrobiology research. techniques revealed many microbiological processes and opened the door to examining the complex chemical interac- MICROBIAL BIODIVERSITY tions of microbial physiology with redox-active minerals (e.g., Newman and Banfield, 2002). In 1994, the Karst Waters Insti- All three domains of life (Eukarya, Bacteria, and Archaea), as tute sponsored the “Breakthroughs in Redox Geochemistry and well as viruses, occur as microscopic life in caves. From a Geomicrobiology” (referred to as “Breakthroughs” herein) con- purely diversity-centric standpoint, understanding who the mi- ference that brought together karst and non-karst microbial sci- crobes are, and how microbial communities relate to each other entists from different biological and geological subdisciplines. from one system to another, are important avenues for future The cross-fertilization had a profound effect and strong growth ecological and geobiological research. in the study of cave and karst geomicrobiology followed. Culture-based methods are perhaps the most desirable, but ge- Characterizing the diversity of microbes in a habitat is impor- nomic approaches have tremendous potential to explore the ex- tant, but now, geomicrobiologists, microbial ecologists, and tent of biodiversity and to uncover novel diversity. Culturing those interested in the microbial sciences, would like to accom- is the biggest stumbling point for species descriptions because plish at least one of two major tasks to understand the functional ~90-99% of microbes from any given habitat are unculturable. roles and distribution of microbial communities in that habitat: Without significant adaptation of culturing techniques and me- (1) culture as many microbes as possible, especially those with dia, regardless of the potentially high numbers of species in key geoecological functions, even though it is possible that a that habitat, culturing diverse metabolic groups is difficult (e.g., microbe may perform differently in the lab than in its natural Amann et al., 1990; Hugenholtz et al., 1998; Zengler et al., habitat; or (2) characterize single cells using non-invasive, in 2002; Radajewski et al., 2003; Schleifer, 2004; Stevenson et situ methods, such as microscopy or other genomics and pro- al., 2004; Wagner et al., 2006). In particular, culturing prac- teomics approaches (e.g., Cohen, 2001; Fernandez, 2005; Hong tices for cave and karst microbes require attention, as many of et al., 2006; Wagner et al., 2006). Broadly speaking, the field is the habitats are oligotrophic (e.g., Cunningham et al., 1995; much closer to attaining the second task than the first. Based on Barton et al., 2004). Culturing often introduces selective bias the previous and ongoing cave and karst microbiological inves- toward microbes able to grow quickly and which out-compete

38 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Caves and Karst as Model Systems

slow-growing organisms. But, culturing allows for quantifica- 2003 ; Gerič et al., 2004; Margesin et al., 2004; Bates et al., tion of metabolically active microbes and can lead to formal 2006; Ikner et al., 2007). The rate of these studies compared descriptions of species (e.g., Margesin et al., 2004), as well as to investigations using molecular techniques has slowed and is enhanced understanding of microbial byproducts that may be not comparable to the recent expansion of culture-based studies useful in applied research (e.g., pharmaceutical industry, bio- and descriptions of new microbial species from diverse habi- engineered systems) (e.g., Herold et al., 2005). tats, especially from the marine realm.

There have been numerous culture-based studies from cave and We have been able to identify microbes that may be difficult, if karst settings that resulted in the characterization of novel spe- not impossible, to cultivate using advanced genetic techniques. cies and elucidation of the roles that the microbes may play 16S rRNA gene surveys are one of the most common strategies in geochemical and geological processes (e.g., Brigmon et al., used to characterize microbial communities, including from 1994; Mikell et al., 1996; Hose et al., 2000; Vlasceanu et al., various caves and karst (Table 1). With the genetic data, mi- 2000; Engel et al., 2001; Lee et al., 2001 ; Canganella et al., crobial species diversity is evaluated, and dataset completeness 2002; Mulec et al., 2002; Khizhnyak et al., 2003; Laiz et al., and biodiversity indicators have also been determined. Many

Table 1: Summary of molecular investigations from caves and karst, differentiated by general habitat type. Many of these studies document the presence of novel species whose similarity to other known species or environmental isolate genetic sequences is relatively low (80-92% sequence identity).

Frontiers of Karst Research 39

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Caves and Karst as Model Systems

of the molecular studies listed in Table 1 resulted in the exami- physiological traits of the organisms, has been employed by nation of bacterial and archaeal diversity in sulfidic and iron/ some researchers (Spilde et al., 2005). Molecular techniques manganese cave or karst habitats, with only a few studies prob- paired with stable isotope systematics and isotope probing in- ing oligotrophic and/or heterotroph-dominated habitats. Most vestigations are also useful. To date, there have been only a of the studies were also conducted with the primary purpose of few isotopic studies (e.g., Pohlman et al., 1997; Vlasceanu et understanding geomicrobiological or geochemical processes, al., 1997; Humphreys, 1999; Vlasceanu et al., 2000; Engel et with limited attention to the potential source of microbes to the al., 2004a; Hutchens et al., 2004), but future studies should caves, or to biogeography and evolution issues. focus on these types of complementary investigations. More- over, whole gene amplification and metagenomics approaches In all practicality, microbes can not be characterized like meta- are critical to characterizing the metabolism of cave and karst zoans, and classification is complicated by different levels microbes, especially when culturing is not possible or success- of genetic relatedness and the extent of gene flow within and ful. between taxa (i.e. horizontal gene transfer). This has resulted in a limited assessment of microbial biodiversity from caves, MICROBES AND ECOSYSTEM which impacts whether or not the biodiversity of karst should FUNCTION consider microbial “species”. Although difficult to consider (at some level) how microbial species could be defined, it is gen- Most cave ecosystems generally reflect being energy- and erally agreed that a microbial species is a monophyletic and nutrient-limited because of their dependence on allochthonous genomically coherent cluster of individuals with a high degree organic material. As such, the flux of microbes in cave streams of similarity in independent traits (e.g., physiological, chemot- or through epikarstic dripwaters can have a significant impact axonomical, and morphological) (e.g., Amann et al., 1990; on the structure and dynamics of cave food webs (e.g., Laiz Hugenholtz et al., 1998; Schleifer, 2004). Moreover, accord- et al., 1999; Gerič et al., 2004). However, the discovery of ing to Cohen (2001), the traditional approach of species defini- chemolithoautotrophically-based ecosystems at the deep-sea tion may not be particularly applicable to bacteria, and instead hydrothermal vents toppled the dogma that all life on earth was microbes within a species are better defined by ecotype, as a dependent on sunlight. In 1986, Cristian Lascu, Serban Sarbu, “set of strains using the same or similar ecological resources” and their colleagues found the uniquely diverse chemolithoau- (Cohen 2002). Clearly, 16S rRNA-based surveys alone are not totrophically-based ecosystem from the hydrogen sulfide-rich adequate to assess the diversity of microbial communities due (sulfidic) groundwater associated with the Movile Cave, Ro- to the highly conservative nature of the 16S rRNA gene and mania (e.g., Sarbu et al., 1996). This discovery was extremely because microbes with genetically identical 16S rRNA gene se- important to advancing cave and karst ecological and geomi- quences can have quite distinct ecological function (e.g., Cohen crobiological research. Despite understanding the importance 2002; Acinas et al., 2004). Therefore, future work describing of chemolithoautotrophy, and knowing that microbes can con- microbes from caves and karst should explore and emphasize tribute to ecosystem carbon and nutrient loading just by their non-rRNA (i.e. functional) gene diversity to answer fundamen- presence (e.g. Pronk et al., 2006), there have been relatively tal questions relating to biodiversity. Cave and karst microbial few studies that document the occurrence, distribution, diver- communities may be particularly important in evaluating the sity, and metabolic potential of microbial communities as ap- microbial species and ecotype concepts because caves are glob- plied to karst ecosystem function, such as for carbon cycling ally distributed and have remarkably similar physicochemical (e.g., Airoldi et al., 1997; Simon et al., 2003; Engel et al., constraints from cave to cave. 2004a; Gerič et al., 2004). For instance, a rich and abundant food source provided by chemolithoautotrophy may reduce Even though these culture-based and molecular studies have nutritional stress to subsurface fauna because members of the gotten us closer to understanding the full diversity of caves, ecosystem would not have to rely on outside food or energy we still do not understand fully microbial metabolism in natu- (e.g., Howarth, 1993). ral settings. Metabolic activity can only tenuously be implied from phylogeny, especially for novel or uncharacterized groups. MICROBIAL ROLES IN GEOCHEMICAL Therefore, additional molecular techniques can identify meta- AND GEOLOGICAL PROCESSES bolic activities much better (e.g., Radajewski et al., 2003; Wag- ner et al., 2006). An iterative process of gleaning information The joint research efforts of geologists and microbiologists, from phylogenetics studies to guide culturing efforts, followed and the growing number of geomicrobiologists trained in in- by phylogenetic analysis of cultures that are displaying the terdisciplinary fields, have promoted new scientific avenues,

40 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Caves and Karst as Model Systems

including the examination of what minerals or geological/geo- deposits from Lechguilla and Spider Caves, New Mexico. chemical processes could be considered biotic and what was Spilde et al. (2005) also demonstrated that some of the min- abiotic. For instance, biologists have discovered that filamen- eral species identified in those deposits can be reproduced in tous morphologies can be minerals, and geologists are discov- vitro by microbial species inoculated from these environments. ering cellular morphologies in their microscopy investigations There are archaeal sequences retrieved from this system and (e.g., Barton et al., 2001). Important advances have been made from the paleofill in Wind Cave, South Dakota (Chelius and in cave and karst geomicrobiology, but many challenges lie be- Moore, 2004). fore us as we study the role of microorganisms in transforming and cycling nutrients and various compounds. Microbes, and Additional forms of poorly crystalline manganese oxides, hy- specifically chemolithoautotrophs, can have a profound impact droxides, and carbonates (e.g. pyrolusite, romanechite, todor- on geochemical and geological processes in karst, although the okite, and rhodochrosite) have been described from caves extent to which we currently can recognize and understand how (Gradzinski et al., 1995; Onac et al., 1997; Northup et al., microbes are a geochemical and geological force in subsurface 2000). In particular, irregularly shaped crusts of manganese settings is likely underappreciated. It is also not clear the extent flowstone (2 – 20 mm thick) are found in Jaskinia Czarna Cave to which microbes have impacted geologic processes through (Tatra Mountains, Poland). Filaments and globular bodies are time, but caves and karst provide unique and accessible sites to interpreted as bacterial or fungal cells that participated in the explore these interactions. formation of the flowstone.

Detailed accounts of previous creation of the cave and ongoing Nitrate Minerals and Nitrogen Cycling cave and karst geomicrobiological studies are included in Jones (2001), Northup and Lavoie (2001), Barton (2006), Barton and Nitrogen is a limiting nutrient in most environments and a Northup (2007) and Engel (2007), so here we will emphasize greater understanding of how microbes influence nitrogen cy- only major accomplishments. Specifically, the highlighted cling is critical to appreciating not only geochemical and geo- studies demonstrate the progress that is being made in under- biological roles, but also ecosystem functioning in karst (Stern standing the geomicrobiological and biogeochemical roles of et al., 2003). At the 1994 “Breakthroughs” conference, George microorganisms in metal and nutrient cycles, including iron Moore wrote “When will we have an accepted explanation for and manganese, nitrogen, sulfur and carbon. As such, advances cave nitrate deposits?” This statement captures the debate that pertaining to carbonate precipitation and dissolution are also has spanned more than a century. Despite the significance of being accomplished. cave nitrates in the early history of caves (i.e. for saltpeter deposits ) there has been little work that has advanced our under- Iron and Manganese Mineral Formation and standing and no consenss has been reached concerning the role Transformations of microbes during nitrate formation in caves. Studies evaluating microbial transformations of nitrogen compounds and There has been considerable progress made in understanding directly testing mineral precipitation reactions and processes the role of microorganisms in iron and manganese cycling in from cave and karst habitats, have not been done. caves. In 1986, Peck (1986) suggested that chemolithoautotrophic microbes were primary producers in the iron and man- Specifically, researches still deliberate over the degree to which ganese deposits in which they were found. Cunningham and bacteria, such as Nitrosomonas spp. and Nitrobacter spp., par- his collaborators were the first to recognize an association be- ticipate in the creation of the cave saltpeter deposits, although tween microbial species and ferromanganese deposits within Nitrobacter spp. had been cultured from nitrogen-rich cave Lechuguilla Cave in New Mexico (Cunningham et al., 1995). sediments in Mammoth Cave (Fliermans et al., 1974). A variety of microbial groups were associated with the ferrihy- drite stalactites, and microbes enhanced precipitation rates up Sulfur Cycling and Transformations to four orders of magnitude compared to inorganic processes (Kasama and Murakami, 2001). Northup et al. (2003) estab- There is a class of caves and karst that developed from waters lished the presence of a diverse community of microorgan- rich in reduced sulfur compounds (mostly hydrogen sulfide), isms, some of whom were related to known manganese and although sulfidic caves are generally rare (e.g., Palmer, 1991). iron-oxidizing bacteria and others who appear to be previously Recently Engel (2007) and Por (2007) reviewed the hydrogeo- unknown. They also documented the presence of metabolically logic origin and ecology of these types of karst systems. Be- active bacteria in the punk rock underlying the ferromanganese cause many sulfur cycle transformations are catalyzed almost

Frontiers of Karst Research 41

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Caves and Karst as Model Systems

exclusively by microbes, sulfidic caves are important to study knowledge of the structure and formation of secondary carbon- because of their rich biodiversity, variety of ecosystem function ate deposits in caves (e.g., Léveillé et al,. 2000; Sanchez-Moral pathways, and unique hydrogeological processes that provide et al., 2003; Barton and Northup, 2007). Numerous studies of insight into groundwater and hydrocarbon reservoir develop- the petrographic fabrics of carbonate deposits coupled to stable ment and evolution. These systems can also be used to proxy isotope ratio analyses have been done to identify microfossils biogeochemical processes in Earth’s history, as the speciation and microbially mediated minerals (e.g., Boston et al., 2001; of sulfur compounds in microbial mats from active sulfidic Melim et al., 2001). Castanier and colleagues (1999) suggested caves was examined to understand sulfur isotope ratio system- that bacterial autotrophic processes cause depletion of carbon atics of ancient geologic materials (Engel et al., 2007). dioxide around cells, favoring carbonate precipitation. From enrichment culturing of microbes from speleothems in Cervo Previous research describing the microbiology of sulfidic caves Cave in Italy, Cacchio et al. (2004) demonstrated using oxygen and stratified systems, like cenotes, sinkholes, and anchialine and carbon isotope ratio systematics that microbes were precip- (or anchihaline) caves, was based on microscopy or culturing itating carbonate at unusually high rates compared to microbes (e.g., Hubbard et al., 1986; Stoessell et al., 1993; Brigmon et from non-karst environments. By using a combination of pe- al., 1994). But, genetic methods have significantly expanded trographic analyses, cultivation, and molecular phylogenetics, the known microbial diversity from these systems (Table 1), Cañaveras and collaborators (1999; 2006) showed that moon- and isotope systematics studies also unveiled possible micro- milk contained numerous filamentous species of Proteobacte- bial metabolic pathways and ecosystem function (e.g., Sarbu et ria that caused calcite to precipitate after colonizing rock sur- al., 1996; Pohlman et al., 1997; Humphreys, 1999; Vlasceanu faces. Several studies have tried to link moonmilk formation et al., 2000; Engel et al., 2004a; Hutchens et al., 2004; Herbert to microbial processes (e.g., Gradzinski et al., 1997; Mulec et et al., 2005). Recent work in the Movile Cave, the Frasassi al., 2002; Galan, 2006), although some researchers found that caves in Italy, and Lower Kane Cave in Wyoming suggests microbes apparently did not play a role during formation in that chemolithoautotrophic sulfur-oxidizing bacteria generate some caves (Borsato et al., 2000). Recently, our knowledge energy that can sustain complex cave ecosystems (e.g., Sarbu of carbonate precipitation by microbes and our experimental et al., 1996; Sarbu et al., 2000; Vlasceanu et al., 2000; Engel sophistication is deepening due to experiments involving cul- et al., 2004a). Observational studies and phylogenetic analy- tured microorganisms, including Bacillus spp. that mediated ses of 16S rRNA genes from microbial mats in these and other calcite precipitation under well-constrained laboratory condi- sulfidic caves indicate that similar microbial lineages are preva- tions (Baskar et al., 2006) and the characterization of genes in lent, including Thiothrix spp. and novel, presently unculturable Bacillus subtilis that are involved in calcite biomineralization Epsilonproteobacteria (e.g., Angert et al., 1998; Engel et al., (Barabesi et al., 2007). 2001; Engel et al., 2003; Engel et al., 2004a; Barton and Lu- iszer, 2005; Engel, 2007; Macalady et al., 2006). Future work Carbonate Dissolution should address the potential biogeographic significance of these microbes in caves and the subsurface as a whole. The classic, textbook, model for karst development involves carbonic acid dissolution, usually at shallow depths near the Compared to neutral pH, carbonate-dominated, subaerial sur- water table. However, carbonate rocks can also dissolve from faces in caves without sulfidic waters, the walls and subaerial reactions with other acids, including microbially produced or- surfaces of sulfidic caves have been the foci of research be- ganic acids and sulfuric acid (e.g., Adkins et al., 1992; White, cause these surfaces often have cave-wall biofilms (snottites) 1997; Perry et al., 2004). The sulfuric acid speleogenesis with extremely acidic pH (0-4) (e.g., Hose et al., 2000; Vl- model was proposed by Stephen Egemeier from his observa- asceanu et al., 2000; Macalady et al., 2007). Interestingly, the tions of Lower Kane Cave (1981). Nearly all of the subsequent

microbial diversity of these biofilms, regardless of the cave, investigations of sulfuric acid speleogenesis assumed that H2S is similar phylogenetically and these extremophile microbial was oxidized to sulfuric acid abiotically, although the presence groups are exceedingly important to speleogenesis (Macalady of microbes in these systems was alluded to have an influence et al., 2007). (e.g., Ball and Jones, 1990; Hill, 1990; Hill, 1995; Sarbu et al., 1996; Angert et al., 1998; Hose et al., 2000; Northup et al., Carbonate Mineral Formation 2000; Sarbu et al., 2000; Vlasceanu et al., 2000; Engel et al., 2001). The role of microbes in speleogenesis was reexamined Understanding biological induction and control on the forma- in Lower Kane Cave, and the sulfur-oxidizing bacteria con-

tion of travertine and carbonate deposits has expanded our sumed nearly all of the H2S in the cave waters, focused acidity

42 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Caves and Karst as Model Systems

on carbonate surfaces, and caused colonization-promoted dis- marine cave with sulphur water springs: Water, Air, and Soil solution (Engel et al., 2004b). Research in other caves and sul- Pollution, v. 99, p. 353-362. fidic aquifers, including the Edwards Aquifer in Central Texas, Amann, R.I., Krumholz, L., and Stahl, D.A., 1990, Fluorescent- has lead to a better understanding of the potential role of mi- oligonucleotide probing of whole cells for determinative, phy- croorganisms during karstification and hydrocarbon reservoir logenetic, and environmental studies in microbiology: Journal modification (Barton and Luiszer, 2005; Macaladyet al., 2006; of Bacteriology, v. 172, p. 762-770. Randall, 2006; Macalady et al., 2007). Amend, J.P., and Teske, A., 2005, Expanding frontiers in deep APPLIED ISSUES SURROUNDING subsurface microbiology Palaeogeography, Palaeoclimatology, CAVE- AND KARST-RELATED Palaeoecology, v. 219, p. 131-155. GEOMICROBIOLOGICAL STUDIES Angert, E.R., Northup, D.E., Reysenbach, A.L., Peek, A.S., Goebel, B.M., and Pace, N.R., 1998, Molecular phylogenetic The presence and activity of microbes in caves and karst may analysis of a bacterial community in Sulphur River, Parker yield beneficial byproducts, such as antimicrobial agents, sur- Cave, Kentucky: American Mineralogist, v. 83, p. 1583-1592. factants, and enzymes, that can be used in applied sciences. Because limited studies have been done, this avenue of re- Ball, T.K., and Jones, J.C., 1990, Speleogenesis in the lime- search has a promising future. Studies by Larry Mallory and stone outcrop north of the South Wales coalfield: the role of colleagues in the 1990s demonstrated that microorganisms cul- micro-organisms in the oxidation of sulphides and hydrocarbons: Cave Science, v. 17, p. 3-8. tured from Hawaiian lava tubes and pools in remote areas of Lechuguilla Cave and Mammoth Cave in Kentucky produced Barabesi, C., Galizzi, A., Mastromei, G., Rossi, M., Tamburini, secondary metabolites that could effectively kill cancer cells M., and Perito, B., 2007, Bacillus subtilis gene cluster involved (L. Mallory, personal communication). An antifungal sub- in calcium carbonate biomineralization: Journal of Bacteriol- stance was produced from Bacillus licheniformis A12, which ogy, v. 189, p. 228-235. was isolated from a cave (Gálvez et al., 1993). Streptomyces tendae, isolated from Grotta dei Cervi in Italy, produced a nov- Barton, H., and Luiszer, F., 2005, Microbial metabolic structure el antibiotic complex, cervimycin, that may help fight multi- in a sulfidic cave hot spring: Potential mechanisms of biospel- eogenesis: Journal of Cave and Karst Studies, v. 67, p. 28-38. drug-resistant pathogens (Herold et al., 2005). The activity of the antibiotic altramiramycin, from a bacterium isolated from Barton, H.A., Spear, J.R., and Pace, N.R., 2001, Microbial life Altramira Cave in Spain, has also been tested (Anonymous, in the underworld: Biogenicity in secondary mineral forma- 2003). Some investigators of moonmilk speculate that this tions: Geomicrobiology Journal, v. 18, p. 1-10. substance was applied to wounds so that patients could benefit from moonmilk’s putative antimicrobial properties (Shaw, Barton, H.A., Taylor, M.R., and Pace, N.R., 2004, Molecular 1997). Because limited studies have been done, this avenue of phylogenetic analysis of a bacterial community in an oligotrophic cave environment: Geomicrobiology Journal, v. 21, p. research has a promising future. 11-20.

REFERENCES Barton, H.A., 2006, Introduction to cave microbiology: A review for the non-specialist: Journal of Cave and Karst Studies, Anonymous, 2003, Biocultural fervour, RTD Info: Magazine v. 68, p. 43-54. on European Research, v. 39, p. 38-39. Barton, H.A., Taylor, M.R., Lubbers, B.R., and Pemberton, Acinas, S.G., Klepac-Ceraj, V., Hunt, D.E., Pharino, C., Ceraj, A.C., 2006, DNA extraction from low biomass carbonate rock: I., Distel, D.L., and Polz, M.F., 2004, Fine-scale phylogenetic An improved method with reduce contamination and the low architecture of a complex bacterial community: Nature, v. 430, biomass contaminant database: Journal of Microbiological p. 551-554. Methods, v. 66, p. 21-31.

Adkins, J.P., Tanner, R.S., Udegbunam, E.O., McInerney, M.J., Barton, H.A., and Northup, D.E., 2007, Geomicrobiology in and Knapp, R.M., 1992, Microbially enhanced oil recovery cave environments: Past, current and future perspectives: Jour- from unconsolidated limestone cores: Geomicrobiology Jour- nal of Cave and Karst Studies, v. 69, p. 163-178. nal, v. 10, p. 77-86. Barton, H.S., Taylor, N.M., Kreate, M.P., Springer, A.C., Oeh- Airoldi, L., Southward, A.J., Niccolai, I., and Cinelli, F., 1997, rle, SA, and Bertog, J.L., 2007, The impact of host rock geo- Sources and pathways of particulate organic carbon in a sub- microbiology on bacterial community structure in oligotropic

Frontiers of Karst Research 43

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Caves and Karst as Model Systems

cave environments: International Journal of Speleology, v. 36, Caumartin, V., 1963, Review of the microbiology of under- p. 93-104. ground environments: Bulletin of the National Speleological Society, v. 25, p. 1-14. Baskar, S., Baskar, R., Mauclaire, L., and McKenzie, J.A., 2006, Microbially induced calcite precipitation in culture ex- Chelius, M.K., and Moore, J.C., 2004, Molecular phylogenetic periments: Possible origin for stalactites in Sahastradhara analysis of Archaea and Bacteria in Wind Cave, South Dakota: caves, Dehradun, India: Current Science, v. 90, p. 58-64. Geomicrobiology Journal, v. 21, p. 123-134.

Bates, C.L., Forstner, M.R.J., Barnes, M.B., Whiteley, M., and Cohen, F.M., 2001, Bacterial species and speciation: System- McLean, R.J.C.,2006, Heterotropic limestone-adherent biofilm atic Biology, v. 50, p. 513-524. isolates from Edwards Aquifer, Texas: The Southwestern Natu- ralist, v. 51, p. 299-309. Cohen, F.M., 2002, What are bacterial species? : Annual Re- views in Microbiology, v. 56, p. 457-487. Borsato, A., Frisia, S., Jones, B., and Van Der Borg, K., 2000, Calcite moonmilk: crystal morphology and environment of Culver, D.C., and Sket, B., 2000, Hotspots of subterranean bio- formation in caves in the Italian Alps: Journal of Sedimentary diversity in caves and wells: Journal of Cave and Karst Studies, Research, v. 70, p. 1171-1182. v. 62, p. 11-17.

Boston, P.J., Spilde, M.N., Northup, D.E., Melim, L.A., Soroka, Cunningham, K.I., Northup, D.E., Pollastro, R.M., Wright, D.S., Kleina, L.G., Lavoie, K.H., Hose, L.D., Mallory, L.M., W.G., and LaRock, E.J., 1995, Bacteria, fungi and biokarst in Dahm, C.N., Crossey, L.J., and Schelble, R.T., 2001, Cave bio- Lechuguilla Cave, Carlsbad Caverns National Park, New Mex- signature suites: Microbes, minerals and Mars: Astrobiology ico: Environmental Geology, v. 25, p. 2-8. Journal, v. 1, p. 25-55. Curtis, T.P., Sloan, W.T., and Scannell, J.W., 2002, Estimating Brigmon, R.L., Martin, H.W., Morris, T.L., Britton, G., and prokaryotic diversity and its limits: Proceedings of the National Zam, S.G., 1994, Biogeochemical ecology of Thiothrix spp. in Academy of Sciences (-USA), v. 99, p. 10494-10499.. underwater limestone caves.: Geomicrobiology Journal, v. 12, p. 141-159. Dykhuizen, D.E., 1998, Santa Rosalia revisited: why are there so many species of bacteria?: Antonie van Leeuwenhoek, v. 73, Cacchio, P., Contento, R., Ercole, C., Cappuccio, G., Preit Mar- p. 25-33. tinez, M., and Lepidi, A., 2004, Involvement of microorganisms in the formation of carbonate speleothems in the Cervo Egemeier, S.J., 1981, Cavern development by thermal waters: Cave (L’Aquila-Italy): Geomicrobiology Journal, v. 21, p. 497- National Speleological Society Bulletin, v. 43, p. 31-51. 509. Engel, A.S., Porter, M.L., Kinkle, B.K., and Kane, T.C., 2001, Cañaveras, J.C., Hoyos, M., Sanchez-Moral, S., Sanz-Rubio, Ecological assessment and geological significance of microbial E., Bedoya, J., Soler, V., Groth, I., Schumann, P., Laiz, L., Gon- communities from Cesspool Cave, Virginia: Geomicrobiology zalez, I., and Saiz-Jimenez, C., 1999, Microbial communities Journal, v. 18, p. 259-274. associated with hydromagnesite and needle-fiber aragonite deposits in a karstic cave (Altamira, Northern Spain): Geomicro- Engel, A.S., Lee, N., Porter, M.L., Stern, L.A., Bennett, P.C., biology Journal, v. 16, p. 9-25. and Wagner, M., 2003, Filamentous "Epsilonproteobacteria" dominate microbial mats from sulfidic cave springs: Applied Cañaveras, J.C., Cuezva, S., Sanchez-Moral, S., Lario, J., Laiz, and Environmental Microbiology, v. 69, p. 5503-5511. L., Gonzalez, J.M., and Saiz-Jimenez, C., 2006, On the origin of fiber calcite crystals in moonmilk deposits: Naturwissen- Engel, A.S., Porter, M.L., Stern, L.A., Quinlan, S., and Ben- schaften, v. 93, p. 27-32. nett, P.C., 2004a, Bacterial diversity and ecosystem function of filamentous microbial mats from aphotic (cave) sulfidic springs Canganella, F., Bianconi, G., Gambacorta, A., Kato, C., and dominated by chemolithoautotrophic "Epsilonproteobacteria": Uematsu, K., 2002, Characterisation of heterotrophic microor- FEMS Microbiology Ecology, v. 51, p. 31-53 ganisms isolated from the "Grotta Azzura" of Cape Palinuro (Salerno, Italy): Marine Ecology, v. 23, p. 1-10. Engel, A.S., Stern, L.A., and Bennett, P.C., 2004b, Microbial contributions to cave formation: new insight into sulfuric acid Castanier, S., LeMetayer-Levrel, G., and Perthuisot, J.-P., 1999, speleogenesis: Geology, v. 32, p. 369-372. Ca-carbonates precipitation and limestone genesis - The micro- biogeologist's point of view: Sedimentary Geology, v. 126, p. Engel, A.S., 2007, On the biodiversity of sulfidic karst habitats: 9-23. Journal of Cave and Karst Studies, v. 69, p. 187-206.

44 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Caves and Karst as Model Systems

Engel, A.S., Lichtenberg, H., Prange, A., and Hormes, J., 2007, Herold, K., Gollmick, F.A., Groth, I., Roth, M., Menzel, K.-D., Speciation of sulfur from filamentous microbial mats from sul- Möllmann, U., Gräfe, U., and Hertweck, C., 2005, Cervimycin fidic cave springs using X-ray absorption near edge spectros- A-D: A polyketide glycoside complex from a cave bacterium copy: FEMS Microbiology Letters, v. 269, p. 54-62. can defeat vancomycin resistance: Chemistry - A European Journal, v. 11, p. 5523 - 5530. Fernandez, L.A., 2005, Exploring prokaryotic diversity: there are other molecular worlds: Molecular Microbiology, v. 55, p. Hill, C., 1995, Sulfur redox reactions - hydrocarbons, native 5-15. sulfur, Mississippi Valley-Type deposits, and sulfuric-acid karst in the Delaware Basin, New-Mexico and Texas: Environmental Fliermans, C.B., Bohlool, B.B., and Schmidt, E.L., 1974 Aute- Geology, v. 25, p. 16-23. cological study of the chemoautotroph Nitrobacter by immunofluorescence: Applied Microbiology, v. 27, p. 124-129. Hill, C.A., 1990, Sulfuric acid speleogenesis of Carlsbad Cav- ern and its relationship to hydrocarbons, Delaware Basin, New Ford, D., and Williams, P., 2007, Karst Hydrogeology and Geo- Mexico and Texas: American Association of Petroleum Geolo- morphology: West Sussex, John Wiley and Sons, Ltd., 562 p. gists Bulletin, v. 74, p. 1685-1694.

Galan, C., 2006, Fauna cavernícola y poblaciones bacteriales Holmes, A.J., Tujula, N.A., Holley, M., Contos, A., James, J.M., de la sima y río subterráneo de mondmilch de Alzola (Gipuz- Rogers, P., and Gillings, M.R., 2001, Phylogenetic structure of koa), in http://www.aranzadi-zientziak.org/fileadmin/docs/es- unusual aquatic microbial formations in Nullarbor caves, Aus- peleologia/AlzolaBiolTr.pdf. tralia: Environmental Microbiology, v. 3, p. 256-264.

Gale, S.J., 1992, Long-term landscape evolution in Australia: Hong , S.E., Bunge, J., Jeon, S.-O., and Epstein, S.S., 2006, Earth Surface Processes and Landforms v. 17, p. 323-343. Predicting microbial species richness: Proceedings of the Na- tional Academy of Science, v. 103, p. 117-122 Gálvez, A., Maqueda, M., Martínez-Bueno, M., Lebbadi, M., and Valdivia, E., 1993, Isolation and physico-chemical charac- Hose, L.D., Palmer, A.N., Palmer, M.V., Northup, D.E., Boston, terization of an antifungal and antibacterial peptide produced P.J., and DuChene, H.R., 2000, Microbiology and geochemistry in a hydrogen-sulphide rich karst environment: Chemical by Bacillus licheniformis A12: Applied Microbiology and Bio- Geology, v. 169, p. 399-423. technology, v. 39, p. 438-442. Howarth, F.G., 1993, High-stress subterranean habitats and Gerič, B., Pipan, T., and Mulec, J., 2004, Diversity of cultur- evolutionary change in cave-inhabiting arthropods: The Ameri- able bacteria and meiofauna in the epikarst of Škocjanske jame can Naturalist, v. 142, p. S65-S77. caves (Slovenia): Acta Carsologica v. 33, p. 301-309. Hubbard, D.A., Herman, J.S., and Bell, P.E., 1986, The role Gonzalez, J.M., Portillo, M.C., and Saiz-Jimenez, C., 2006, of sulfide oxidation in the genesis of Cesspool Cave, Virginia, Metabolically-active Crenarchaeota in Altamira Cave: Natur- wissenschaften, v. 93, p. 42-45. USA, in 9th International Congress of Speleology, Volume 1: Barcelona, Spain, p. 255-257. Gradzinski, M., Szulc, J., and Smyk, B., 1997, Microbial agents of moonmilk calcification, in Jeannin, P.-Y., ed., Proceedings of Hugenholtz, P., Goebel, B.M., and Pace, N.R., 1998, Impact the 12th International Congress of Speleology, Volume 1, Inter- of culture-independent studies on the emerging phylogenetic national Union of Speleology, p. 275-278. view of bacterial diversity: Journal of Bacteriology, v. 180, p. 4765-4774. Gradziński, M., Banaś, M., and Uchman, A., 1995, Biogenic origin of manganese flowstones from Jaskinia Czarna Cave, Humphreys, W.F., 1999, Physico-chemical profile and energy Tatra Mts., Western Carpathians: Annales Societatis Geologo- fixation in Bundera Sinkhole, an anchialine remiped habitat in rum Poloniae, v. 65, p. 19-27. north-western Australia: Journal of the Royal Society of West- ern Australia, v. 82, p. 89-98. Groth, I. and Saiz-Jimenez, C., 1999, Actinomycetes in hypo- Hutchens, E., Radajewski, S., Dumont, M.G., McDonald, I.R., gean environments: Geomicrobiology Journal, v. 16, p. 1-8. and Murrell, J.C., 2004, Analysis of methanotrophic bacteria in Movile Cave by stable isotope probing: Environmental Micro- Herbert, R.A., Ranchou-Peyruse, A., Duran, R., Guyoneaud, biology, v. 6, p. 111-120. R., and Schwabe, S., 2005, Characterization of purple sulfur bacteria from the South Andros Black Hole cave system: high- Ikner, L.A., Toomey, R.S., Nolan, G., Neilson, J.W., Pryor, lights taxonomic problems for ecological studies among the B.M., and Maier, R.M., 2007, Culturable microbial diversity genera Allochromatium and Thiocapsa: Environmental Micro- and impact of tourism in Kartchner Caverns, Arizona: Micro- biology, v. 7, 1260-1268. bial Ecology, v. 53, p. 30-42.

Frontiers of Karst Research 45

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Caves and Karst as Model Systems

Jones, B., 2001, Microbial activity in caves: a geological per- Meisinger, D.B., Zimmermann, J., Ludwig, W., Schleifer, K.- spective: Geomicrobiology Journal, v. 18, p. 345-358. H., Wanner, G., Schmid, M., Bennett, P.C., Engel, A.S., and Lee, N.M., 2007, In situ detection of novel Acidobacteria in Kasama, T., and Murakami, T., 2001, The effect of microorgan- microbial mats from a chemolithoautotrophically based cave isms on Fe precipitation rates at neutral pH: Chem Geol, v. 180, ecosystem (Lower Kane Cave, WY, USA): Environmental Mi- p. 117-128. crobiology, v. 9, p. 1523–1534.

Khizhnyak, S.V., Tausheva, I.V., Berezikova, A.A., Nesteren- Melim, L.A., Shinglman, K.M., Boston, P.J., Northup, D.E., ko, E.V., and Rogozin, D.Y., 2003, Psychrophilic and psychro- Splide, M.N., and Queen, J.M., 2001, Evidence of microbial tolerant heterotrophic microorganisms of middle Siberian karst involvement in pool finger precipitation, Hidden Cave, New cavities: Russian Journal of Ecology, v. 34, p. 231-235. Mexico: Geomicrobiology Journal v. 18, p. 311-330.

Laiz, L., Gonzalez-Delvalle, M., Hermosin, B., Ortiz-Martinez, Mikell, A.T., Smith, C.L., and Richardson, J.C., 1996, Evalu- A., and Saiz-Jimenez, C., 2003 Isolation of cave bacteria and ation of media and techniques to enumerate heterotrophic mi- substrate utilization at different temperatures: Geomicrobiol- crobes from karst and sand aquifer springs: Microbial Ecology, ogy Journal, v. 20, p. 479-489. v. 31, p. 115-124.

Laiz, L., Groth, I., Gonzalez, I., Saiz-Jimenez, C., 1999, Micro- Mulec, J., Zalar, P., Zupan–Hajna, N., and Rupnik, M., 2002, biological study of the dripping waters in Altamira Cave (San- Screening for culturable microorganisms from cave environ- tillana del Mar, Spain): Journal of Microbiological Methods, v. ments (Slovenia): Acta Carsologica, v. 31, p. 177-187. 36, p. 129-138. Newman, D.K., and Banfield, J.F., 2002, Geomicrobiology: Lee, S.D., Kim, E.S., Roe, J.H., Kim, J.H., Kang, S.O., and How molecular-scale interactions underpin biogeochemical Hah, Y.C., 2000, Saccharothrix violacea sp. nov., isolated from systems: Science, v. 296, p. 1071-1077. a gold mine cave, and Saccharothrix albidocapillata comb. nov.: International Journal of Systematic and Evolutionary Mi- Northup, D.E., Dahm, C.N., Melim, L.A., Spilde, M.N., crobiology, v. 50, p. 1315-1323. Crossey, L.J., Lavoie, K.H., Mallory, L.M., Boston, P.J., Cun- ningham, K.I., and Barns, S.M., 2000, Evidence for geomicro- Lee, S.D., Kim, E.S., Min, K.L., Lee, W.Y., Kang, S.O., and biological interactions in Guadalupe caves: Journal of Cave Hah, Y.C., 2001, Pseudonocardia kongjuensis sp. nov., isolated and Karst Studies, v. 62, p. 80-90. from a gold mine cave: International Journal of Systematic and Evolutionary Microbiology, v. 51, p. 1505-1510. Northup, D.E., and Lavoie, K., 2001, Geomicrobiology of caves: a review: Geomicrobiology Journal, v. 18, p. 199-222. Léveillé, R.J., Fyfe, W.S. and Longstaffe, F.J;, 2000, Geomi- crobiology of carbonate-silicate microbalites from Hawaiian Northup, D.E., Barns, S.M., Yu, L.E., Spilde, M.N., Schelble, basaltic sea caves: Chemical Geology, v. 169, p. 339-355. R.T., Dano, K.E., Crossey, L.J., Connolly, C.A., Boston, P.J., Natvig, D.O., and Dahm, C.N., 2003, Diverse microbial com- Macalady, J.L., Jones, D.S., and Lyon, E.H., 2007, Extremely munities inhabiting ferromanganese deposits in Lechuguilla acidic, pendulous cave wall biofilms from the Frasassi cave sys- and Spider Caves: Environmental Microbiology, v. 5, p. 1071- tem, Italy: Environmental Microbiology, v. 9, p. 1402-1414. 1086.

Macalady, J.L.L., Lyon, E.H., Koffman, B., Albertson, L.K., Northup, D.E., Connolly, C.A., Trent, A., Boston, P.J., Peck, Meyer, K., Galdenzi, S., and Mariani, S., 2006, Dominant mi- V.M., and Natvig, D.O., 2008, On the nature of bacterial com- crobial populations in limestone-corroding stream biofilms, munities from Four Windows Cave, El Malpais National Mon- Frasassi cave system, Italy: Applied and Environmental Micro- ument, New Mexico, USA: AMCS Bulletin, v. 19, p. 81. biology, v. 72, p. 5596-5609. Onac, B.P., Tysseland, M., Bengeanu, M., and Hofenpradli, A., Mahler, B.J., Personné, J.-C., Lods, G.F., and Drogue, C., 2000, 1997, Deposition of black manganese and iron-rich sediments Transport of free and particulate-associated bacteria in karst: in Vântului Cave (Romania), in Jeannin, P.-Y., ed., Proceedings Journal of Hydrology, v. 238, p. 179-193. of the 12th International Congress of Speleology, Volume 1, International Union of Speleology, p. 235-238. Margesin, R., Schumann, P., Spröer, C., and Gounot, A.-M., Pace, N.R., 1997, A molecular view of microbial diversity and 2004, Arthrobacter psychrophenolicus sp. nov., isolated from the biosphere: Science, v. 276, p. 734-740. an alpine ice cave: International Journal of Systematic and Evolutionary Microbiology, v. 54, p. 2067-2072. Palmer, A.N., 1991, Origin and morphology of limestone caves: Geological Society of America Bulletin, v. 103, p. 1-21.

46 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Caves and Karst as Model Systems

Palmer, A.N., 1995, Geochemical models for the origin of Schabereiter-Gurtner, C., Saiz-Jimenez, C., Pinar, G., Lubitz, macroscopic solution porosity in carbonate rocks, in Budd, A., W., and Rolleke, S., 2002a, Phylogenetic 16S rRNA analysis Saller, A., and Harris, P., eds., Unconformities and Porosity in reveals the presence of complex and partly unknown bacterial Carbonate Strata: Memoir 63: Tulsa, AAPG, p. 77-101. communities in Tito Bustillo Cave, Spain, and on its Palaeoli- thic paintings: Environmental Microbiology, v. 4, p. 392-400. Peck, S.B., 1986, Bacterial deposition of iron and manganese oxides in North American caves: National Speleological Soci- Schabereiter-Gurtner, C., Saiz-Jimenez, C., Piñar, G., Lubitz, ety Bulletin v. 48, p. 26-30. W., and Rölleke, S., 2002b, Altamira Cave Paleolithic paintings harbour partly unknown bacterial communities: FEMS Micro- Perry, T.D., Duckworth, O.W., McNamara, C.J., Martin, S.T., biology Letters, v. 211, p. 7-11. and Mitchell, R., 2004, Effects of the biologically produced polymer alginic acid on macroscopic and microscopic calcite Schabereiter-Gurtner, C., Saiz-Jimenez, C., Piñar, G., Lubitz, dissolution rates: Environmental Science and Technology v. 38, W., and Rölleke, S., 2004, Phylogenetic diversity of bacteria p. 3040-3046. associated with Paleolithic paintings and surrounding rock walls in two Spanish caves (Llonín and La Garma): FEMS Mi- Pohlman, J.W., Iliffe, T.M., and Cifuentes, L.A., 1997, A stable crobiology Ecology, v. 47, p. 235-247. isotope study of organic cycling and the ecology of an anchialine cave ecosystem: Marine Ecology Progress Series, v. 155, Schleifer, K.H., 2004, Microbial diversity: Facts, problems and p. 17-27. prospects: Systematic and Applied Microbiology, v. 27, p. 3-9.

Por, F.D., 2007, Ophel: a groundwater biome based on chemoau- Shaw, T.R., 1997, Historical introduction, in Hill, C.A., and totrophic resources. The global significance of the Ayyalon Forti, P., eds., Cave Minerals of the World: Huntsville, AL, Na- cave finds, Israel: Hydrobiologia, v. 592, p. 1-10. tional Speleological Society, p. 27-43.

Pronk, M., Goldscheider, N., and Zopfi, J., 2006, Dynamics and Simon, K., Benfield, E.F., and Macko, S.A., 2003, Food web interaction of organic carbon, turbidity and bacteria in a karst structure and the role of epilithic biofilms in cave streams: aquifer system: Hydrogeology Journal, v. 14, p. 473-484. Ecology, v. 84, p. 2395-2406.

Radajewski, S., McDonald, I.R., and Murrell, J.C., 2003, Sta- Spilde, M.N., Northup, D.E., Boston, P.J., Schelble, R.T., Dano, ble-isotope probing of nucleic acids: a window to the function K.E., Crossey, L.J., and Dahm, C.N., 2005, Geomicrobiology of uncultured microorganisms: Current Opinion in Biotechnol- of cave ferromanganese deposits: A field and laboratory inves- ogy, v. 14, p. 296-302. tigation: Geomicrobiology Journal, v. 22, p. 99-116.

Randall, K.W., 2006, Assessing the potential impact of mi- Stern, L.A., Engel, A.S., and Bennett, P.C., 2003, Nitrogen crobes in the Edwards and Trinity Aquifers of Central Texas isotope evidence of ammonia vapor assimilation by cave wall [MS thesis], Louisiana State University. microbial biofilms in a sulfidic cave, a novel mechanism of nutrient acquisition, American Geophysical Union, Fall Meeting Rusterholtz, K., and Mallory, L.M., 1994, Density, activity and 2003, abstract #B42E-05. diversity of bacteria indigenous to a karstic aquifer: Microbial Ecology, v. 28, p. 79-99. Stevens, T., 1997, Lithoautotrophy in the subsurface: FEMS Microbiology Reviews, v. 20, p. 327-337. Sanchez-Moral, S., Cañaveras, J.C., Laiz, L., Saiz-Jimenez, C., Bedoya, J., and Luque, L., 2003, Biomediated precipitation Stevenson, B.S., Eichorst, S.A., Wertz, J.T., Schmidt, T.M., and of calcium carbonate metastable phases in hypogean environ- Breznak, J.A., 2004, New strategies for cultivation and detec- ments: A short review: Geomicrobiology Journal, v. 20, p. 491 tion of previously uncultured microbes: Applied and Environ- - 500. mental Microbiology, v. 70, p. 4748-4755.

Sarbu, S.M., Kane, T.C., and Kinkle, B.K., 1996, A chemoau- Stoessell, R.K., Moore, Y.H., and Coke, J.G., 1993, The oc- totrophically based cave ecosystem: Science, v. 272, p. 1953- currence and effect of sulfate reduction and sulfide oxidation 1955. on coastal limestone dissolution in Yucatan cenotes: Ground Water, v. 31, p. 566-575. Sarbu, S.M., Galdenzi, S., Menichetti, M., and Gentile, G., 2000, Geology and biology of Grotte di Frasassi (Frasassi Symk, B., and Drzal, M., 1964, Research on the influence Caves) in Central Italy, an ecological multi-disciplinary study of microorganisms on the development of karst phenomena: of a hypogenic underground karst system, in Wilkens, H., Cul- Geographia Polonica, v. 2, p. 57-60. ver, D., and Humphreys, S., eds., Ecosystems of the World: Subterranean Ecosystems, Volume 30: Oxford, Elsevier Sci- ence, p. 361-381.

Frontiers of Karst Research 47

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Caves and Karst as Model Systems

Taylor, M.R., 1999, Dark Life: Martian Nanobacteria, Rock- Austrian Central Alps, and evidence of ammonia-oxidizing eating Cave Bugs, and Other Extreme organisms of Inner Earth Crenarchaeota source: Applied and Environmental Microbiol- and Outer Space: New York, New York, Scribner, 287 p. ogy, v. 73, p. 259-270.

van Beynen, P., and Townsend, K., 2005, A disturbance index White, W.B., 1988, Geomorphology and Hydrology of Karst for karst environments: Environmental Management, v. 36, p. Terrains: New York, Oxford University Press, 464 p. 101-116. White, W.B., 1997, Thermodynamic equilibrium, kinetics, acti- Vlasceanu, L., Popa, R., and Kinkle, B., 1997, Characteriza- vation barriers, and reaction mechanisms for chemical reactions tion of Thiobacillus thioparus LV43 and its distribution in a in Karst Terrains: Environmental Geology, v. 30, p. 46-58. chemoautotrophically based groundwater ecosystem: Applied and Environmental Microbiology, v. 63, p. 3123-3127. Whitman, W.B., Coleman, D.C. and Wiebe, W.J., 1998, Prokaryotes: The unseen majority: Proceedings of the National Vlasceanu, L., Sarbu, S.M., Engel, A.S., and Kinkle, B.K., Academy of Science, v. 95, p. 6578-6583. 2000, Acidic cave-wall biofilms located in the Frasassi Gorge, Italy: Geomicrobiology Journal, v. 17, p. 125-139. Zengler, K., Toledo, G., Rappe, M., Elkins, J., Mathur, E.J., Short, J.M., and Keller, M., 2002, Cultivating the uncultured: Wagner, M., Nielsen, P.H., Loy, A., Nielsen, J.L., and Daims, Proceedings of the National Academy of Sciences, v. 99, p. H., 2006, Linking microbial community structure with func- 15681-15686. tion: fluorescence in situ hybridization -microautoradiography and isotope arrays: Current Opinion in Biotechnology, v. 17, Zimmermann, J., Gonzalez, J.M., Saiz-Jimenez, C., and Lud- p. 83-91. wig, W., 2005, Detection and phylogenetic relationships of highly diverse uncultured acidobacterial communities in Alta- Weidler, G.W., Dornmayr-Pfaffenhuemer, M., Gerbl, F.W., Hei- mira Cave using 23S rRNA sequence analyses: Geomicrobiol- nen, W., and Stan-Lotter, H., 2007, Communities of Archaea ogy Journal, v. 22, p. 379-388. and Bacteria in a subsurface radioactive thermal spring in the

48 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Ecosystem Science and Karst Systems

Ecosystem Science and Karst Systems

Kevin S. Simon School of Biology and Ecology University of Maine Orono, Maine, 04469-5722

INTRODUCTION • Energy limitation in karst • Nutrient biogeochemistry cosystem science focuses on two major functional attributes of ecological systems: energy flux and nutrient APPROPRIATE SPATIAL SCALES FOR Ebiogeochemistry. Historically, the study of these eco- EXAMINING ECOSYSTEM FUNCTION logical functions in karst has lagged study in other fields such as biodiversity, evolution, and hydrology. This lag is not due The field of ecosystem ecology was revolutionized by linking to the insignificance of ecosystem functions in karst, rather it is abiotic and biotic systems at large scales (i.e. an ecosystem). most likely a result of the relative youth of the field of ecosys- The classic example is the use of a watershed as a unit of study tem ecology, which did not begin to fully emerge until 1960’s (Bormann and Likens, 1967). This approach allowed for the and 70’s. However, the pace of advances in ecosystem science construction of input/output budgets and subsequent large- in karst has not matched that focused on other ecosystems on scale experimental work. Raymond Rouch and his colleagues the surface. This is ironic considering that perhaps the first and pioneered the idea of using a karst basin as an ecosystem, akin best known example of ecosystem ecology was done in a karst to the use of a watershed on the surface. In a series of more ecosystem; that is the development of an energy budget for Sil- than 20 papers (summarized in Rouch, 1986) Rouch measured ver Spring by Eugene Odum (1957). the inputs and outputs of animals by sampling springs draining different portions of the Baget Basin. Rouch also integrated Why study ecosystem function in karst? First, energy avail- geological and hydrological thinking by examining subcompo- ability is considered a key driver of the ecology and evolution nents of the basin, specifically the epikarst, unsaturated, and of cave communities. Caves have long been thought to be saturated zones. This integration of the physical sciences into energy-limited because of the absence of photosynthesis and ecology is a hallmark of ecosystem science that has not been the features of cave animals such as reduced metabolic rate, fully exploited in karst. larger but fewer eggs, and increased longevity are considered to support this assertion (see review by Hűppop, 2000). Second, Gibert (1986), in what is the first true ecosystem study in karst, karst ecosystems have much to offer to the study of ecosystem used Rouch’s karst basin framework and quantified the flux of science in general. Some karst ecosystems are exclusively het- organic carbon from springs draining the epikarst and the satu- erotrophic while others are fueled by internal chemolithoauto- rated zones of the Dorvan-Cleyzieu basin in France. Among trophy and both represent endpoints in the broad spectrum of Gibert’s most important findings were that dissolved organic ecosystem types. The study of extremes often yields insights carbon represented a larger flux than particulate organic carbon, into the more common. Third, future changes related to human that those fluxes were spatially and temporally variable, and activity such as global climate change and intensifying human that microbes were likely to be key players in mediating energy landuse are likely to be manifest in karst through alterations to transfer between organic carbon and animals in karst. The use ecosystem function. For example, changes to surface vegeta- of a karst basin as an ecosystem is an important approach, but tion and hydrology, both likely outcomes of climate and lan- quantifying ecosystem processes at that scale is made difficult duse change, will undoubtedly influence energy flux in karst. by the lack of accessibility to much of karst basins (e.g. the epikarst). This is one reason that we have seen little progress What is the current state of ecosystem science in subterranean beyond that of Rouch and Gibert on studying ecosystem func- habitats? In this brief review, I consider: tion at the basin scale.

• Appropriate scales for examining ecosystem function The decade following the work of Gibert was relatively un- • Foodweb structure and energy sources eventful, until new work emerged that focused on a smaller • Spatial and temporal variation in ecosystem function spatial scale, the stream reach, which integrated the thinking

Frontiers of Karst Research 49

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Ecosystem Science and Karst Systems

and methods of stream ecology into the study of karst (e.g. karst are dissolved and particulate organic matter. While dis- Graening and Brown, 2003, Simon et al., 2003). Work in this solved organic matter has been recognized as a potential food area has expanded our knowledge of food web structure, en- (Gibert, 1986), particulate matter such as leaves, wood and ergy sources, and organic matter dynamics in the unsaturated guano has typically been thought of as very important foods zone of karst. This finer-scale approach has not yet been ap- in “food-poor” caves. Somewhat surprisingly, the stable iso- plied to the epikarst and saturated zones of aquifers, prob- tope data for the four detritus-based cave streams all suggested ably because of the difficulty in accessing and sampling these that leaves, wood and bat guano were largely unused by most subsystems. Virtually all of the research done at any spatial cave animals (Graening and Brown, 2003; Simon et al., 2003). scale in karst has been descriptive or comparative. To date, Rather, microbial films on rocks (epilithon) and fine sediments no ecosystem-scale experimental manipulations, such as the were much more important foods for all animals. An experi- catchment logging at Hubbard Brook (Bormann and Likens, mental addition of a stable isotope tracer that labeled microbial 1967), have been attempted. At the reach scale, only a few films further confirmed that energy fixed in microbial biomass small-scale manipulations of organic matter or nutrients have was incorporated in grazers and, ultimately predators (Simon et been attempted. (e.g. Simon and Benfield, 2001). This lack of al., 2003). The most likely source of energy fueling those mi- an experimental approach is perhaps the most glaring weakness crobial films, based on natural abundance of15 N, was dissolved of the work done on ecosystem function in karst. organic matter originating from soils and percolating through the epikarst. Why cave animals rely mostly on microbial films FOOD-WEB STRUCTURE AND ENERGY and dissolved organic matter may be related to the relative re- SOURCES liability of those foods compared to leaves and wood which are patchily distributed, but the true answer is unknown. One We now have a much better picture of what karst food webs caveat to the results of these studies is that they represent snap- look like and a better understanding of what organic matter shots in time and indicate only foods that were important over sources fuel these food webs than we did only a decade ago perhaps the previous months. Energy sources such as leaves (Fig. 1). This advance is largely due to the use of stable iso- and wood may be quite important in brief, periodic pulses and/ topes. The natural abundances of stable isotopes, especially or in restricted locations in basins. 13C and 15N, are now widely used tools for untangling foodweb structure and sorting out the important energy sources to SPATIAL AND TEMPORAL VARIATION IN foodwebs. This approach has been used to delineate a chemo- ECOSYSTEM FUNCTION lithoautotroph-based foodweb (Sarbu et al., 1996) and the detritus-based foodwebs of 4 cave streams (Graening and Brown, Ecosystem function is likely to vary considerably over space 2003; Simon et al., 2003). Other, partial food chains also have and time considering the heterogeneity of aquifer structure and been elucidated using stable isotopes (e.g. Opsahl and Chanton, temporal variability in hydrology. Both of these factors likely 2006). The chemolithoautotroph-based foodweb, both terres- contribute to variation in the availability of organic matter and trial and aquatic, comprised 3 trophic levels including bacteria, physical conditions (water velocity, temperature) that can influ- invertebrate grazers and predators. ence ecosystem function. Within karst basins, organic matter input and availability is both spatially and temporally variable. The 4 detritus-based cave streams were also comprised of 3 Gibert (1986) and Simon et al. (2001) found that concentrations trophic levels: detritus and associated microbes, invertebrate of dissolved organic carbon were higher in the saturated zone grazers and invertebrate and vertebrate predators. There ap- than in the epikarst of the Dorvan basin in France. The amounts pears to be no quantitative data about food web structure or of organic matter present in streams varies within aquifers (Si- energy use for the terrestrial foodweb in detritus-based karst mon and Benfield, 2001) and is linked, in part, to the presence systems. Beyond trophic levels, there remains considerable of openings to the surface that permit entry of coarse organic ambiguity in the specific trophic interactions among species. matter such as leaves and wood. There can also be substantial In fact, we know surprisingly little about what individual taxa spatial variation in decomposition rates. In Organ Cave, USA, can and do eat in caves. Stable isotope tracers (e.g. Simon et for example, variation in leaf breakdown rates among cave al., 2003), analyses of gut contents, and feeding trials (e.g. Si- streams spans the range of breakdown rates observed among mon and Buikema, 1997) should provide further resolution of surface streams (Simon and Benfield, 2001). That spatial vari- trophic interactions in karst food webs. ation was largely explained by differences in the invertebrate communities, particularly the abundance of Gammarus minus, The primary energy sources to most aquatic communities in a leaf-shredding amphipod that is a stygophile. The distribu-

50 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Ecosystem Science and Karst Systems

Figure 1. Generalized aquatic foodwebs in karst. Arrows indicate energy flux and arrow size indicates presumed importance of each flux based on existing data. CPOM = coarse particulate organic matter, FPOM = fine particulate organic matter, DOM = dissolved organic matter. tion of G. minus in Organ Cave, and therefore the spatial pat- Temporal variation in organic matter and nutrient availability tern of leaf breakdown, apparently is the result of the historical can be considerable in karst and is strongly influenced by vari- changes in aquifer structure that influenced G. minus invasion ability in hydrologic inputs. Concentrations of organic carbon, into the cave (Culver et al., 1995). Aside from these few ex- nitrate and microbial concentration in the water column all re- amples, we have little understanding of the extent of spatial spond to flood pulses. In some cases concentrations of DOC variation in energy or nutrient input or what controls that spa- and bacteria are positively correlated with discharge, but not tial variation. Such spatial variability may be quite important necessarily in all locations within a basin (Pronk et al., 2006, in regulating community structure. For example, at a broad Simon et al., 2001). Many parameters demonstrate a clear tem- geographic scale, subterranean biodiversity corresponds to sur- poral lag in their relation to flood pulses. For example, Pronk face productivity (Culver et al., 2006). Whether there are “hot et al. (2006) found peaks in bacterial density and concentra- spots” or “hot times” of energy input and processing that drive tions of dissolved organic carbon and nitrate that lagged peaks community diversity within karst basins is unknown. in discharge by several days in a Swiss aquifer. These lags in

Frontiers of Karst Research 51

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Ecosystem Science and Karst Systems

response are probably due at least in part to differences in travel Engel and Northup). However, there are very few measures of times of water through the various flowpaths through soils and nutrient (e.g. N, P, S) uptake, retention, transport and transfor- the aquifer. Linking hydrological information to measures of mation in karst. In many karst systems the availability of N organic matter and nutrient availability (e.g. Pronk et al., 2006) and P may be relatively unimportant because they are generally will be critical for developing conceptual and predictive mod- abundant. For example, in Organ Cave, USA ammonium up- els of both spatial and temporal variation in ecosystem function take in streams was relatively low compared to surface systems in karst. and the availability of N and P did not limit microbial biomass or respiration on wood substrates (Simon and Benfield, 2002). ENERGY LIMITATION IN KARST In this case, the concept of nutrient spiraling was a useful tool for studying biogeochemistry in the stream flow component of The idea that karst foodwebs are energy limited is firmly en- a karst aquifer. Different techniques and conceptual models trenched despite the lack of any experimental evidence. An- will be needed for other subsystems such as the epikarst and ecdotal and indirect evidence both support and refute the idea saturated zones in karst. In chemolithoautotrophic systems, the of energy limitation. Anecdotally, cave animals can be quite availability of inorganic elements (especially C, N, S and Fe) abundant where organic matter is abundant in caves (e.g. old may be critical to limitation of biological productivity. wooden structures) and where organic pollution occurs (e.g. Si- mon and Buikema, 1997). Some measures of ecosystem func- ACKNOWLEDGEMENTS tion, particularly stream metabolism and leaf decomposition, shed some light on the topic. Compared to surface streams, This summary was substantially improved by discussions with the amount of organic matter in cave streams is low (Simon Daniel Fong, Lara Hinderstein, Bridget Maloney, Robert Payn, and Benfield, 2002), except where organic pollution occurs Michael Venarsky, and Frank Wilhelm. (e.g. Graening and Brown, 2003). Despite low abundance of organic matter, the rate of cave stream respiration can be REFERENCES disproportionately high, resulting in high specific respiration and rapid organic matter turnover times (Simon and Benfield, Bormann, F.H., and Likens, G.E., 1967, Nutrient cycling: Sci- 2002). This efficient and rapid use of organic matter is con- ence, v. 155, p. 424-429. sistent with energy limitation. Rates of leaf breakdown have Brown A.V. and Schram, M.D., 1982, Leaf detritus processing been measured in at least four different studies (Brussock et al., in an Ozark cave stream: Arkansas Academy of Science Pro- 1988; Brown and Schram, 1982; Galas et al., 1996; Simon and ceedings, v. 36, p. 14-16. Benfield, 2001). If energy is highly limiting, one would expect leaves to be rapidly consumed and leaf breakdown rates to be Brussock, P.P., Willis, L.D. and A.V. Brown, A.V., 1988, Leaf high. However, in most of the caves studied, leaf breakdown decomposition in an Ozark cave and spring: Journal of Fresh- was quite slow relative to surface streams, possibly related to water Ecology, v. 4, p. 263-269 an observed lack of animals that could consume, or “shred”, the Culver, D.C., Kane, T.C. and Fong, D.W., 1995. Adaptation and leaves that enter caves. In other experimental results, higher Natural Selection in Caves: The Evolution of Gammarus mi- leaf breakdown rates were observed in cave streams. In Or- nus: Cambridge, Harvard University Press, 223p. gan Cave, USA, leaf breakdown rates were very rapid, but only where Gammarus minus, a stygophile, was abundant. It may Culver, D.C., Deharveng, L., Bedos, A., Lewis, J.J., Madden, be the case that energy limitation varies within and among karst M., Reddell, J.R., Sket, B., Trontelj, P., and White, D.,2006, basins. The issue of energy limitation will remain unresolved The mid-latitude biodiversity ridge in terrestrial cave fauna: until energy availability is experimentally manipulated in a Ecography, v. 29, p. 120-128. karst system. Galas, J., Bednarz, T., Dumnicka, E., Starzecka, A. and Wojtan, K., 1996, Litter decomposition in a mountain cave water: Ar- NUTRIENT BIOGEOCHEMISTRY chiv für Hydrobiologie, v. 138, p. 199-211.

Compared to studies of organic matter, less is known about bio- Gibert, J., 1986, Ecologie d'un systeme karstique jurassien. Hy- geochemical function in karst. The interactions of microbes in drogéologie, dérive animale, transits de matières, dynamique geochemistry, for example cave formations, has received much de la population de Niphargus (Crustacé Amphipode): Mem- recent attention and our understanding of chemolithoautrophic oires de Biospeologie, v. 13, p.1-379. ecosystems is rapidly expanding (see the report by Summers-

52 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Ecosystem Science and Karst Systems

Graening, G.O. and Brown, A.V., 2003, Ecosystem dynamics Simon, K.S. and Buikema, A.L. Jr., 1997, Effects of organic and pollution effects in an Ozark cave stream: Journal of the pollution on an Appalachian cave: changes in macroinverte- American Water Resources Association v. 39, p. 497-505. brate populations and food supplies: The American Midland Naturalist, v. 138, p. 387-401. Hűppop, K., 2000, How do cave animals cope with the food scarcity in caves?: in Subterranean Ecosystems, H. Wilkens, Simon, K.S., Gibert, J., Petitot, P. and Laurent, R., 2001, Spa- D.C. Culver and W.F. Humphreys [Eds.], Amsterdam, Elsevier, tial and temporal patterns of bacterial density and metabolic pp. 159-188. activity in a karst aquifer: Archiv für Hydrobiologie, v. 151, p. 67-82. Odum, H.T., 1957, Trophic structure and productivity of Silver Springs, Florida: Ecological Monographs, v. 27, p. 55-112. Simon, K.S., and Benfield, E.F., 2001, Leaf and wood breakdown in cave streams: Journal of the North American Bentho- Opsahl, S. P. and Chanton, J.P., 2006, Isotopic evidence for logical Society, v. 482, p. 31-39. methane-based chemosynthesis in the upper Floridan aquifer food web: Oecologia, v. 150, p. 89-96. Simon, K.S., and Benfield, E.F., 2002, Ammomium retention and whole stream metabolism in cave streams: Hydrobiologia, Pronk, M., Goldscheider, N., Zopfi, J., 2006, Dynamics of or- v. 482, p. 31-39. ganic carbon, turbidity, and bacteria in a karst aquifer system: Hydrogeology Journal, v. 14, p. 473-484. Simon, K.S., Benfield, E.F., and Macko, S.A., 2003, Food web structure and the role of epilithic films in cave streams: Ecol- Rouch, R., 1986, Sur l’ecologie des eaux souterraines dans la ogy, v. 84, p. 2395-2406. karst: Stygologia, v. 2, p. 352-398.

Sarbu, S.M., Kane, T.C., and Kinkel, B.F., 1996, A chemoautotrophically based cave ecosystem. Science, v. 272, p. 1953- 1955

Frontiers of Karst Research 53

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 The Struggle to Measure Subterranean Biodiversity

The Struggle to Measure Subterranean Biodiversity

David C. Culver Department of Biology American University 4400 Massachusetts Ave. NW Washington DC 20016

INTRODUCTION Zagmajster (Slovenia). The second key factor was the multi- national study of European aquatic subterranean biodiversity, n spite of the considerable interest in caves as ecological, Protocols for the Assessment and Conservation of Aquatic Life evolutionary, and microbiological laboratories (see Poul- In the Subsurface (PASCALIS). A preliminary summary of son and White, 1969; Culver et al., 1994; Engel et al., this work is in Gibert (2005), and a more complete analysis is I forthcoming as a special issue of the journal Freshwater Biol- 2004), the heart of speleobiology remains the description and explanation of species diversity. While the study of biodiver- ogy. sity is important in any habitat, it seems especially so for subterranean habitats. There are undoubtedly several reasons for SPECIES DESCRIPTION AND this but certainly the often bizarre morphology of subterranean IDENTIFICATION animals combined with the seemingly inexhaustible supply of undescribed species is a major component. There appears to be a perception in the United States that species description and identification of cave animals is on the What is the current state of biodiversity studies in subterranean decline. It is certainly true that a remarkable group of taxono- habitats? In this brief review, I will take a broad view of subter- mists, including Barr, Bowman, Christiansen, Ferguson, Hobbs ranean biodiversity, and consider Jr., Hoffman, Holsinger, and Muchmore, taxonomists who largely defined the field in the U.S., are deceased, retired, or • Species description and identification close to retirement. Nevertheless, species description contin- • Sampling strategies and completeness ues at a healthy pace, including monographic revisions (e.g., • Biodiversity mapping Zhang and Holsinger, 2003; Miller, 2005) and species descrip- • Explanations of biodiversity patterns tions (e.g., Lewis, 2005, Lewis et al., 2006). This is not to say that there are not groups on which no taxonomist is currently For the first topic (species description and identification), I will working. These groups include planarians and carabid beetles. focus on the situation in the United States, but for the others It is interesting to note that for both of these groups there have I will take a more international perspective. An international been relatively recent revisions (Barr, 2004; Kenk, 1977), and perspective is not only appropriate but essential because many so this makes them less attractive to the systematists looking of the recent advances come from outside the U.S., especially for major groups to revise. This is likely to become a grow- Europe. ing problem of finding someone to either identify or describe species from relatively well known groups, such as the carabid If this review had been written ten years ago, there would have beetle genus Pseudanophthalmus. Novel solutions are needed. been very little to say about the last two points, and not that Even training many more taxonomists is unlikely to solve this much to say about sampling completeness. In the last ten years problem since taxonomists tend to work on groups in need of there has been a remarkable increase in quantitative approaches major revision, rather than minor additions. to sampling and biodiversity mapping. This has been the result of many factors including the usual increases in computer Another trend has been the internationalization of the taxono- capacity and capability. There are two other key factors worth my of subterranean organisms. For example, the Czech scien- noting. The first is the emerging generation of speleobiologists tist Zacharda (1985) and the French scientist Thibaud (1996) with strong quantitative skills and a focus on subterranean bio- are actively interested in American cave mites and Collembola diversity. Among the most notable of these are Stefan Eberhard respectively. In addition, the Slovenian scientist Tanja Pipan (Australia), David Ferreira (France), Tristan Lefébure (France), is collaborating with Janet Reid on the description of cave co- Tanja Pipan (Slovenia), Katie Schneider (U.S.A.) and Maja pepods.

54 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 The Struggle to Measure Subterranean Biodiversity

Another intriguing trend is represented by Julian J. Lewis in formation available, but given that new caves and other karst the U.S., Boris Sket in Slovenia, and in an earlier generation, sites are continually being discovered, it will never be possible Claude Delamare Deboutteville in France. They all moved be- to generalize if one waits for completeness. More relevant is yond the standard idea that one taxonomist can only specialize how complete sampling is. This may seem to be an impossible on one group to do significant taxonomic work on disparate question to answer but in fact, there are many techniques avail- groups of stygobionts and troglobionts. For example, Lewis able to answer this very question, many of them available in the has described both isopods and millipedes (e.g., Lewis, 2005; freeware Estimates (Colwell, 2005). Lewis et al., 2006). Two general types of estimates have been used by speleobiolo- A final trend worth noting is the role (or more accurately the gists. The first relies on a species accumulation curve produced lack of a role) that molecular data plays in the taxonomy of sub- by randomizing caves (or quadrats) and selecting 1,2,3… sites terranean organisms. There have been numerous studies that at random and re-sampling 100 or more times. For the most have shown that cryptic species are quite common. Among part, such accumulation curves have not reached an asymptote the most striking examples are from the European fauna, sum- (see Schneider and Culver, 2004 for one of the first such stud- marized by Trontelj et al. (2007). They found that all species ies in caves and Rouch and Danielopol, 1997 for an example examined with ranges of more than about 250 km in maximum from a hyporheic habitat). Exceptions to this have been either linear extent actually consisted of two or more cryptic species. very intensive sampling at small scales such as epikarst drips One of their examples is the iconic European cave salamander within a cave (Pipan and Culver, 2007) or intensively sampled Proteus anguinus, which actually comprises at least three cryp- areas aggregated into equal-sized sampling areas (Culver et al., tic species (see also Gorički and Trontelj, 2006). A good North 2006). The second relies on estimators of the total number of American example, albeit based on protein rather than DNA species, typically based on some variant of the ratio of species variation is that of Kane et al. (1992) on the carabid beetle found in a single site relative to those found in two sites. In Neaphaenops tellkampfi. What is remarkable is that, in spite several cases, these estimates are on the order of magnitude of overwhelming molecular evidence, none of these cryptic of 50 percent higher than the number of known species (e.g., species have been formally described. This would afford tax- Pipan and Culver, 2007). onomists of subterranean groups the opportunity to set a new standard. What remains to be determined is which of these estimators, e.g, Chao2, bootstrap, ICE, etc. are most appropriate for cave SAMPLING STRATEGIES AND data. What makes cave data unique is that unlike most samples COMPLETENESS in surface environments, the number of species known from a single site does not decline to zero as is typical of samples of There is probably not even a single relatively species-rich cave, surface-dwelling fauna, but rather reaches a nonzero asymp- let alone a region for which we can be confident that all cave- tote. This makes biological sense because of the high level of limited species (aquatic stygobionts and terrestrial troglobionts) endemism in caves (see Christman et al., 2005). In relatively have been discovered and described. In the two most species- low diversity cave regions, such as the Walloon karst in Bel- rich caves known and among the best studied—Vjetrenica in gium, levels of endemism are lower and the number of species Bosnia & Hercegovina and the Postojna – Planina Cave System known from a single site does approach zero (Deharveng et in Slovenia—new species are being discovered (I. Lučič, pers. al., in press). Whether it makes statistical sense to use these comm., S. Polak pers. comm.). The most recent issue of the estimators for the cave fauna is still an open question, requir- journal Subterranean Biology (volume 4) nicely illustrates this ing a more detailed examination of the assumptions used in the point. Among the articles are a description of a 200 km range various estimators. extension of a stygobiotic isopod in France (Notenboom et al., 2006), a new genus of beetle from Albanian caves (Giachino There are other approaches to the question of sampling com- and Vailati, 2006), and three new species of stygobiotic amphi- pleteness. One idea is adaptive sampling (Christman, 2004), pods from Mediterranean islands (Messouli et al., 2006). which is a sampling strategy that relies on prior knowledge of where organisms are and modifies sampling accordingly. So, While there is a tendency on the part of some speleobiolo- for example, species rich cave areas would be sampled more in- gists to conclude that no generalizations are possible because tensively than species poor areas, which is of course analogous sampling is incomplete, this approach is simplistic at best and to the frequent claim of sampling bias. It is important to note generally wrong. Of course, we all need to have the best in-

Frontiers of Karst Research 55

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 The Struggle to Measure Subterranean Biodiversity

that species rich cave regions repay additional sampling effort mapping has become much more informative and predictive. more than species poor regions, a prediction made by Zagma- Nearly all investigators who have mapped subterranean species jster (2007) and by Deharveng et al. (in press) in a summary richness have found that it is very heterogeneous, and this cre- of the PASCALIS project. Another approach has been to take ates difficulties on both displaying and interpreting a map. Het- historical snapshots of data to determine how much the pattern erogeneity is greatly reduced if data are grouped into quadrats. has changed with time (Culver et al., 2004a; Ferreira, 2005). In This largely eliminates the problem of poorly sampled caves general the conclusion has been that additional data have not and well sampled caves since the data are lumped together (see changed the basic pattern. Culver et al., 2004a).

All of these statistical and historical analyses assume that sampling at sites is more or less complete and that the sites The problem of heterogeneity brings about the question of what themselves are a representative sample of subterranean habitats scale of analysis is optimal, i.e., what is the optimal quadrat in karst. The work of Meštrov (1962) on superficial subter- size. Zagmajster et al. (2008) suggest that the optimal size is ranean seeps (hypotelminorheic), Gers (1983) on superficial when Moran’s I (a measure of spatial autocorrelation) reach- subterranean terrestrial habitats (milieu souterrain superficiel es its maximum, and apply this to the rich troglobiotic beetle or M.S.S.), expanded the range of subterranean habitats out- fauna of the Balkans. Zagmajster (2007) has also produced side karst areas, as did earlier and ongoing work on freshwater contour maps of species richness based on both observed num- interstitial habitats (e.g., Lewis and Holsinger, 1985). More bers of species and predicted number based on estimators such importantly for determining diversity patterns on karst is the as Chao’s. This allows an analysis of where it is likely most work on the fauna of epikarst drips, most extensively stud- uncollected species are, and it turns out that in her case, it is in ied by Pipan (2005). The rich copepod fauna of epikarst can the peaks of richness rather than the valleys. double the known number of stygobionts in a cave (Pipan and Culver, 2005), especially in North America where subterranean Another approach is that of Christman (2005) who very ele- copepods are understudied. Major discoveries of this kind ob- gantly partitioned stygobiotic species richness in counties in viously cannot be accounted for by the statistical models of the southeastern United States into that predicted by number of species richness. caves in the county (local effects), a spatial autocorrelation me- soscale component, and residual variation. It not only displays In general, the quality of sampling of subterranean biodiver- the pattern of species richness, it also partitions it into local sity has seen major advances in the past decade due to several and regional components. Christman also points out the diffi- factors. The first of these, and one that is exemplified by the cult problem of modeling the spatial distribution because of the PASCALIS project, is a shift from individual work to collab- large number of zeroes, zeroes that may be because no habitats orative work, involving specialists of many taxonomic groups, exist in the county, habitats exist but have not been sampled, an ABTI-like approach. The second is to incorporate more ef- and habitats exist, have been sampled, but no stygobionts were ficient and complete sampling strategies. This includes target- found. This problem of “zero inflation” is very typical of any ing undersampled regions (like southeast Asia), undersampled large scale mapping of subterranean biodiversity and the prob- habitats (like epikarst), and targeting undersampled groups lem is far from solved. (like copepods). The third is the internationalization of cave biodiversity studies. This may involve specialists from more Finally, nearly all maps of subterranean biodiversity show not and more countries involved in the same projects, and the emer- only excess zeroes, but very clear hotspots of species richness, gence of new countries and new researchers, such as China and hotspots that we cannot always predict (but see Culver et al., Chinese researchers. Finally, there has been a big increase in 2004b). This has been observed by Zagmajster et al. (2008) the links between biodiversity measurement and conservation for the Dinaric beetle fauna, by Deharveng and Latella (unpub- issues, and in fact most agencies that fund biodiversity studies lished) for the Chinese beetle fauna, Deharveng et al. (2007) require this link. for the European stygofauna, and Christman et al. (2005) for the single cave terrestrial endemics of eastern North America. BIODIVERSITY MAPPING EXPLANATIONS OF BIODIVERSITY With the widespread availability of sophisticated mapping PATTERNS software such as ArcMap© and the development of spatial statistical methods such as conditional autoregression (CAR), For decades there has been an interest in subterranean biodi-

56 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 The Struggle to Measure Subterranean Biodiversity

versity patterns, initially focused on tropical-temperate com- Christman, M.C., 2004, Sequential sampling for rare or geo- parisons, which were all the rage in ecology especially in the graphically clustered populations: in W.L. Thompson [ed.], 1970’s. The initial pattern of higher diversity in temperate ar- Sampling rare or elusive species, Washington, DC, Island eas was disputed (e.g. Howarth, 1972) but this discussion was Press, p. 134-145. framed in the absence of quantitative data. It was not until Christman, M.C., 2005, Mapping subterranean biodiversity: Deharveng’s (2005) summary of species richness in some well- In D.C. Culver and W.B. White [eds.] Encyclopedia of Caves, studied tropical caves that it was clear that temperate caves Amsterdam, Elsevier/Academic Press, p. 355-361. were more diverse, although this has still not been demonstrated quantitatively on a regional scale. A complicating factor is that Christman, M.C., and Culver, D.C., 2001, The relationship be- faunal composition in terms of ecological groups is extremely tween cave biodiversity and available habitat: Journal of Bio- different in tropical and temperate caves. If guanobionts are geography, v 28, p. 367-380. included (most are only known from caves), it may change the Christman, M.C., Culver, D.C., Madden, M. and White, D., pattern substantially. 2005, Patterns of endemism of the eastern North American cave fauna: Journal of Biogeography, v. 32, p. 1441-1452. In a quantitative study involving data for more than 1,500 caves, Culver et al. (2006) found that troglobiotic diversity Colwell, R.K., 2005, EstimateS: Statistical estimation of spe- in temperate areas was highest in regions of high actual pro- cies richness and shared species from samples Version 7.5, ductivity which formed a band in southern Europe and south- User’s Guide and application published at http://purl.oclc.org/ eastern United States, suggesting that the species richness was estimates. controlled by resource availability. On a broad regional scale, Colwell, R.K., Mao, C.X. and Chang, J., 2004, Interpolating, others have found other factors to be important, including dis- extrapolating, and comparing incidence-based species accumu- tance to recent glaciation (Castellarini et al., 2007), hydraulic lation curves: Ecology, v. 85, p. 2717–2727. conductivity (Castellarini et al., 2007), and habitat availability (cave density) (Christman and Culver, 2001). At a smaller Culver, D.C., Christman, M.C., Sket, B. and Trontelj, P., 2004a, scale, chemical factors can be important, as has been shown by Sampling adequacy in an extreme environment: species rich- Pipan, Blejec, and Brancelj (2006). The multivariate analysis ness patterns in Slovenian caves: Biodiversity and Conserva- tion, v.13, p. 1209-1229. of the causes of differences in subterranean species richness is in its infancy and no clear generalities have yet emerged. In- Culver, D.C., Christman, M.C., Šereg, I., Trontelj, P. and Sket, deed, there is no reason to expect that all cave communities are B., 2004b, The location of terrestrial species-rich caves in a limited by the same set of factors. cave-rich area: Subterranean Biology, v. 2, p. 27-32.

ACKNOWLEDGEMENTS Culver, D.C., Deharveng, L., Bedos, A., Lewis, J.J., Madden, M., Reddell, J.R., Sket, B., Trontelj, P. and White, D., 2006, The mid-latitude biodiversity ridge in terrestrial cave fauna: Many discussions with Anne Bedos, Louis Deharveng, Tanja Ecography, v. 29, p. 120-128. Pipan, and Maja Zagmajster about subterranean biodiversity have guided much of my thinking. Louis Deharveng, Dan- Culver, D.C., Kane, T.C., and Fong, D.W., 1995, Adaptation iel Fong, Janine Gibert, and Tanja Pipan all reviewed various and Natural Selection in Caves: Cambridge, Harvard Univer- drafts of this paper. sity Press, 223 p.

REFERENCES Deharveng, L., 2005, Diversity patterns in the tropics: in D.C. Culver and W.B. White [eds.] Encyclopedia of caves, Amster- Barr, T.C., Jr., 2004, A classification and checklist of the genus dam, Elsevier/Academic Press, p. 166-170. Pseudanophthalmus Jeannel (Coleoptera: Carabidae: Trechi- Deharveng, L., Stoch, F., Gibert, J., Bedos, A., Galassi, D., nae): Virginia Museum of Natural History Special Publication Zagmajster, M., Brancelj A., Camacho A., Fiers, F., Martin, P., 11, 52 p. Giani, N., Magniez, G. and Marmonier, P., in press, Insight into the groundwater biodiversity of western Europe: Freshwater Castellarini, F., Malard, F., Dole-Olivier, M.-J. and Gibert, Biology. J., 2007, Modeling the distribution of stygobionts in the Jura Mountains (eastern France). Implications for the protection of Engel, A.S., Stern, L.A. and Bennett, P.C., 2004, Microbial ground waters: Diversity and Distributions, v. 13, p. 213-224. contributions to cave formation: new insights into sulfuric acid speleogenesis: Geology, v. 32(5), p. 369-372.

Frontiers of Karst Research 57

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 The Struggle to Measure Subterranean Biodiversity

Ferreira, D., 2005, Biodiversite aquatique souterraine de Meštrov, M., 1962, Un nouveau milieu aquatique souterrain: France: base de donnees, patrons de distribution et implications le biotope hypotelminorhéique: Compte Rendus. Academie des en termes de conservation: Ph.D. Thesis, Universite Claude Sciences Paris, v. 254, p. 2677-2679. Bernard-Lyon 1. Miller, J.A., 2005a, Cave adaptation in the spider genus Anthro- Gers, C., 1983, Etude morphologie et biometrique de Speono- bia (Araneae, Liknyphiidae, Erigoninae): Zoologica Scripta, v. mus carrerei Foures, 1953 (Coleopteres Bathysciinae) recolte 34, p. 565-592. dans deux grottes et dans le milieu souterrain superficial: Mem- oires de Biospeologie, v. 10, p. 265-276. Miller, J.A., 2005b, A redescription of Porrhomma cavernicola Keyserling (Araneae, Linyphiidae) with notes on Appalachian Giachino, P.M., and Vailati, D., 2006, Kircheria beroni, a new troglobites: Journal of Arachnology, v. 33, p. 426-438. genus and new species of subterranean hygropericolous Lepto- dirinae from Albania (Coleoptera, Cholevidae): Subterranean Notenboom, J., Oertel, A., Boutin,C. and Deharveng, L., 2006, Biology, v.4, p. 103-116. Range extension of the karst water isopod Sphaeromides ray- mondi (Cirolanidae, Isopoda, Crustacea) in France: Subterra- Gibert, J. [ed.], 2005, World Subterranean Biodiversity: Pro- nean Biology, v. 4, p. 9-14. ceedings of an international symposium held on 8 – 10 Decem- ber in Villeurbanne, France. Equipe Hydrobiologie et Ecolo- Pipan, T., 2005, Epikarst - a promising habitat. Copepod fauna, gie Souterraines, Université Claude Bernard I, Villeurbanne, its diversity and ecology: a case study from Slovenia (Europe): France. ZRC Publishing, Karst Research Institute at ZRC SAZU, Lju- bljana, 100 p. Gorički Š. and Trontelj, P., 2006, Structure and evolution of the mitochondrial control region and flanking sequences in the Pipan, T., Blejec, A. and Brancelj, A., 2006, Multivariate analy- European cave salamander Proteus anguinus: Gene, v. 378, p. sis of copepod assemblages in epikarstic waters of some Slove- 31-41. nian caves: Hydrobiologia, v. 559, p. 213-223.

Howarth, F. G., 1972, Cavernicoles in lava tubes on the island Pipan, T., and Culver, D.C., 2005, Estimating biodiversity in of Hawaii: Science, v. 175, p. 325-326. the epikarstic zone of a West Virginia cave: Journal of Cave and Karst Studies, v. 67, p. 103-109. Kane, T. C., Barr, T.C., Jr. and Badaracca, W., 1992, Cave beetle genetics: geology and gene flow: Heredity, v. 68, p. 277-286. Pipan, T., and Culver, D.C., 2007, Regional species richness in an obligate subterranean dwelling fauna—epikarst copepods: Kenk, R., 1977, Freshwater triclads (Turbellaria) of North Journal of Biogeography, v. 34, p. 854-861. America, IX: the genus Sphalloplana: Smithsonian Contribu- tions to Zoology, no. 246, 38 p. Poulson, T. L. and White, W.B., 1969, The cave environment: Science, v. 165, p. 971-981. Lewis, J. J., 2005, Six new species of Pseudotremia from caves of the Tennessee Cumberland Plateau (Diplopoda: Chordeuma- Rouch, R., and Danielopol, D.L., 1997, Species richness of mi- tida: Cleidogonidae): Zootoxa, v. 1080, p. 17-31. crocrustacea in subterranean freshwater habitats. Comparative analysis and approximate evaluation: International Revue Ges. Lewis J.J., Graening, G.O., Fenolio, D.B., et al., 2006, Cae- Hydrobiologie, v. 82, p. 121-145. cidotea mackini, new species, with a synopsis of the subter- Schneider, K. and Culver, D.C., 2004, Estimating subterra- ranean asellids of Oklahoma (Crustacea : Isopoda : Asellidae): nean species richness using intensive sampling and rarefaction Proceedings of the Biological Society of Washington, v. 119, curves in a high density cave region in West Virginia: Journal p. 563-575. of Cave and Karst Studies, v. 66, p. 39-45. Lewis, J., and Holsinger, J.R., 1985, Caecidotea phreatica, a Thibaud, J.-M., 1996, Une nouvelle espèce de Typhlogastrura new phreatobitic isopod crustacean (Asellidae) from southeastern Virginia: Proceedings of the Biological Society of Wash- de deux grottes des Ètats-Unis d'Amérique (Collembola, Hypo- ington, v. 98, p. 1004-1011. gastrurideae): Revue francais Entomologiique, v. 18, p. 11-12.

Messouli, M., Messana, G. and Yacoubi-Khebiza, M., 2006, Trontelj, P., Douady, C.J., Fiser, C., Gibert, J., Goricki, S., Three new species of Pseudoniphargus (Amphipoda), from the Lefébure, T., Sket, B. and Zaksek, V., 2007, A molecular test groundwater of three Mediterranean islands, with notes on the for cryptic diversity in groundwater: how large are the ranges of macro-stygobionts?: In J. Gibert and D.C. Culver [eds.], Ps. Adriaticus: Subterranean Biology, v. 4, p. 79-102. Freshwater Biology, Online early: doi: 10.1111/j.1365-

58 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 The Struggle to Measure Subterranean Biodiversity

2427.2007.01877.x ness patterns of obligate subterranean beetles in a global biodi- Zacharda, M., 1985, New Rhagidiidae (Acarina: Prostigmata) versity hotspot - effect of scale and sampling intensity: Diver- from caves of the U.S.A.: Věstník Československé Společnosti sity and Distributions, v. 14, p. 95-105. Zoologické, v. 49, p.67-80. Zhang, J., and J.R. Holsinger, J.R., 2003, Systematics of the Zagmajster, M., 2007, Analiza razširjenosti izbranih skupin tro- freshwater amphipod genus Crangonyx (Crangonyctidae) in globiotske favne na Dinarskem območju: Ph.D. Dissertation, North America: Virginia Museum of Natural History, v. 6, p. University of Ljubljana. 1-274.

Zagmajster, M., Culver, D.C. and Sket, B. 2008. Species rich-

Frontiers of Karst Research 59

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Evolution in Karst Lineages

Evolution in Karst Lineages, Ages, and Adaptation of Cave Faunas

Megan Porter Department of Biological Sciences University of Maryland Baltimore County 1000 Hilltop Circle Baltimore, MD 21250

INTRODUCTION • The evolutionary timeline of cave adaptation

he study of subterranean faunas has a long and rich This review will focus on these three general themes, describ- history, with the first scientific description of a cave- ing the current state of research in each. adapted species occurring in the 1700s. A major theme T BIODIVERSITY AND PHYLOGEOGRAPHY in biospeleological studies has been the evolution of troglomorphy – the unique suite of traits characterizing ‘cave-adapted’ species. The suite of changes associated with cave adaptation Since the inception of ‘biospeleology’ as a scientific field, includes reduction in pigment, eye size, and metabolic rates, there has been a strong focus on species descriptions. These and hypertrophy of nonoptic sensory organs and feeding struc- taxonomic studies have fueled biogeographic research aimed tures. From an evolutionary perspective, the troglomorphic at understanding subterranean species diversity and distribu- state is interesting for several reasons. First, the highly conver- tions. Although these issues have been intensively studied for gent form can obscure taxonomic relationships among cave- several hundred years, the application of molecular techniques adapted species and among closely related cave and epigean to appreciating the genetic diversity and biogeography of cave (i.e. surface) species, hindering distributional and diversity fauna have led to several significant discoveries that have large studies. Second, because of the universal convergence of form implications to the way subterranean biodiversity is quantified found in the cave environment, being exhibited in structural, and protected. functional, and behavioral changes across diverse taxonomic groups, cave-adapted species provide a model evolutionary One of the foundations of biogeographic studies is a solid under- system for studying the forces that have lead to such a strong standing of the distribution of the species of interest. The most convergence and for investigating the relationship between ge- striking discovery of recent molecular phylogenetic studies of notypic and phenotypic change. cave faunas is the identification of cryptic genetic diversity, which impacts our understanding of the distribution of subter- Molecular techniques employed in the evolutionary study of ranean faunas and their relationships with each other and with cave species began in the 1970s with the development of the epigean species. Specifically, almost every recent, sequence- first major molecular markers, allozymes (protein variants, based molecular study of cave-adapted species has identified Avise and Selander, 1972; Carmody et al., 1972; Hetrick and at least one cryptic lineage (Buhay and Crandall, 2005; Goricki Gooch, 1973; Laing et al., 1976; Sbordoni et al., 1979). As the and Trontelj, 2006; Lefébure et al., 2006b; Buhay et al., 2007), field of molecular evolution and phylogenetics advanced, the while in some groups cryptic genetic diversity appears to be available number of molecular markers increased and the as- rampant (Finston et al., 2007; Lefébure et al., 2007). In many sociated analyses became more sophisticated. However, only cases, this previously unidentified diversity is diagnosed as the within the last few years have molecular studies using gene se- presence of cryptic species. Even when morphological-based quences to investigate evolutionary questions in cave faunas taxonomy and molecular studies agree on the diagnosis of a become common. cave-adapted species, convergent morphologies often lead to erroneous hypotheses of close evolutionary relationships at These recent molecular studies in subterranean habitats encom- higher taxonomic levels (Wiens et al., 2003). For example, pass several major evolutionary themes: molecular studies of the stygobiotic catfish Prietella phreatophila Carranza, 1954 and Prietella lundbergi Walsh and Gilbert, • Biodiversity and phylogeography 1995 indicate that each is more closely related to species from • The genetic and developmental mechanisms of cave ad- different genera than they are to each other (P. phreatophila to aptation Ictalurus spp. and P. lundbergi to Ameiurus spp. (Wilcox et al.,

60 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Evolution in Karst Lineages

2004). A similar situation exists in cave-adapted Troglocaris 2000; Parra-Olea, 2003; Buhay and Crandall, 2005; Finston et shrimp species, which represents a non-monophyletic assem- al., 2007). blage with several species being most closely related to surface species or different cave-adapted species outside of the genus THE GENETIC AND DEVELOPMENTAL Troglocaris (Zaksek et al., 2007). MECHANISMS OF CAVE ADAPTATION

This situation calls for the marriage of classic morphology- In contrast to their confounding nature in biodiversity stud- based taxonomy with molecular techniques to fully under- ies, the convergent morphologies of the cave-adapted forms stand subterranean biodiversity. One of the best examples of provide a unique system for understanding the genetic and joining these two fields to understand the karst biodiversity is developmental mechanisms responsible for constructive (ac- from a study of cave-adapted spider Cicurina spp. from Tex- quired) and regressive (lost) traits. Perhaps the best-studied, as, USA. This species complex represents some of the most cave-adapted species with respect to understanding the issues common problems with taxonomic studies of cave faunas: low of convergent trait acquisition and loss is the Mexican cave tet- population densities and a rarity of encountering individuals, ra, Astyanax mexicanus. A. mexicanus contains both eyed and particularly the adult specimens required for accurate species pigmented surface populations and eyeless, pigmentless cave- identification and taxonomic description. Using molecular adapted populations (Fig. 1), providing an unrivalled system data, immature specimens can be compared to known species for mechanistic studies of trait evolution. Direct comparisons for determination of how they fit into the current, morpholo- can be made between a derived trait (the cave form) and its an- gy-based, taxonomic scheme. In Texas, this approach has led cestral state (the surface form) within one species. Along with to identification of the federally endangered Cicurina madla this, the numerous regressive and constructive traits associated from more than twice the number of previously reported caves with the cave form and the existence of multiple cave-adapted (Paquin and Hedin, 2004). In a broader effort to bridge the gap populations arising from independent subterranean coloniza- and provide a way to integrate DNA information with current tion events make A. mexicanus a model system for research in taxonomic methods, Lefébure et al. (2006a) empirically evalu- evolutionary biology (Jeffery, 2001). ated the correlation between currently described crustacean species and molecular divergence. Based on sequence data for two mitochondrial genes, they propose a species level molecular threshold to be used as a tool for the taxonomic delimitation of new crustacean species.

In the absence of obvious morphological differences due to Figure 1. Epigean (A) and hypogean (B) forms of Astyanax mexicanus (photos provided by W.R. Jeffery). extreme convergence, molecular phylogenetic studies of cave- adapted species have been successful at diagnosing cryptic genetic diversity and the presence of taxonomic incongruence. Studies of trait evolution in A. mexicanus also have a significant At a minimum, these recent studies illustrate a decoupling be- history in biospeleology, beginning with classical genetic stud- tween molecular and morphological evolution in the cave envi- ies where individuals from different populations were crossed ronment (Lefébure et al., 2006b). If patterns of cryptic genetic to examine the heritability of eye size and pigmentation (Sado- diversity within currently described species and misdiagnosed glu, 1955, 1956, 1958; Wilkens, 1971). More recent genetic relationships above the species level hold for most subterranean studies have used genome-wide linkage maps with quantitative faunas, then taxonomic based studies may be extreme under- trait analysis to investigate the genetic basis of pigment loss estimates of subterranean biodiversity and our understanding (e.g. the evolution of albinism) in two different populations of of basic species-level relationships may be biased (Proudlove cave-adapted A. mexicanus (Protas et al., 2006). This study and Wood, 2003; Lefébure et al., 2006b; Finston et al., 2007). found that albinism in these populations was linked to a known Although there is considerable debate about how to deal with pigmentation gene, Oca2. Furthermore, different inactivating this cryptic genetic diversity within the existing taxonomic mutations were identified, indicating that albinism has evolved framework, understanding these patterns will greatly impact in- independently in the two studied populations by convergent terpretations of hydrological connectivity patterns, ecological evolution in the same gene (Protas et al., 2006). In one of habitat restrictions, geological barriers, climatic and geological the few studies to study the genetic basis of cave-adapted trait vicariant events, and differing invasion scenarios contributing loss outside of A. mexicanus, Leys et al. (2005) investigated to the genetic structuring of populations (Chippindale et al., the eye pigment gene cinnabar in independent lineages of

Frontiers of Karst Research 61

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Evolution in Karst Lineages

subterranean dytiscid water beetles. Similar to the findings of lighted in the studies presented here for just a few cave-adapted Protas et al. (2006), cinnabar sequences indicate increased species hold across other regressive traits within A. mexicanus rates of sequence evolution, including mutations leading to the and within other cave-adapted species. loss of gene function. THE EVOLUTIONARY TIMELINE OF CAVE Loss of eyes and pigmentation has also been investigated in A. ADAPTATION mexicanus from a developmental perspective. In particular, developmental studies of eye reduction and loss show a different An important aspect of understanding troglomorphy is the evo- mechanistic pattern than pigment genes. In cavefish embryos, lutionary time required to achieve the cave form. Because it is expanded midline signaling of the hedgehog genes inhibits eye difficult to pinpoint the time of subterranean colonization and formation, leading to lens apoptosis and eye degeneration (Ya- isolation from the surface, the time of cave adaptation is gener- mamoto et al., 2004). Because hedgehog genes may regulate ally thought of in relative terms, where the degree of eye and the development of constructive traits in cavefish morphologies pigment reduction indicates the period of cavernicolous evolu- such as feeding structures (Jeffery, 2005) and possibly olfac- tion and therefore the relative phylogenetic age of each species tory regions in the brain (Menuet et al., 2007), it has been hy- (Aden, 2005). Advances in molecular methods, however, al- pothesized that loss of eyes in A. mexicanus is a consequence low for the estimation of divergence times among subterranean of the pleiotropic effects of natural selection for constructive lineages. Estimations of divergence times in cave species have traits (Jeffery, 2005). Studies of other cave-adapted traits, such been used most often to correlate the time of lineage splitting as loss of aggressive behaviors, found at least one A. mexicanus with geologic or climatic events that may have caused subterra- population that had not lost aggression (Espinasa et al., 2005). nean invasion, such as Pleistocene glaciations or late Miocene These results indicate that there are differences among traits, aridification. With one exception where pairs of cave-adapted and all cave-associated traits need to be examined to determine Orconectes spp. have estimated divergence times ranging from patterns and mechanisms of acquisition and loss. 102-125 Ma (Buhay and Crandall, 2005; but see Trontelj, 2007 for disagreement with these ages), most estimates of subterra- In the history of biospeleological research, there has been a nean colonization times are younger than ~16 Ma. Strikingly, long-standing debate over whether the convergent loss of traits these estimates fall within similar ranges, even through the es- in the cave-adapted form is due to selective pressures or neutral timates come from different cave-adapted taxonomic groups mutations and drift (Kane and Richardson, 1985). However, and regions of the world (Australian dytiscid beetles, 9.0-15.8 the most recent genetic and developmental studies highlighted Ma - Cooper et al., 2002; Dinaric karst isopods, 0.8-3.9 Ma - here make it clear that this debate is no longer valid. Instead, Verovnik et al., 2004; European amphipods, >13 Ma - Lefébure these studies emphasize that both selection and drift play a role et al., 2006b; Australian crangonyctoid amphipods, 4.1-14.7 in cave-adapted morphologies; at what point along the evolu- Ma - Cooper et al., 2007; Mexican cave tetras, 0.3-5.2 Ma - tionary path to cave adaptation selection and drift act, and the Porter et al., 2007; Dinaric karst decapods, 3.7-5.3 Ma - Zaksek degree to which each is the driving force, depends on the trait. et al., 2007). Quantitative trait loci (QTL) mapping in A. mexicanus illustrate the differences among cave-associated regressed traits. In the efforts to understand how fast (or slow) acquisition of the Using this approach, Protas et al. (2007) showed that cave cave morphology occurs, A. mexicanus again offers a unique alleles at eye or lens QTLs caused size reductions, consistent model. The primary mode of A. mexicanus subterranean colo- with evolution by natural selection while QTLs associated with nization is via stream capture, with most of the captured surface melanophores caused both increases and decreases in number, drainages no longer supporting epigean populations (Mitchell consistent with drift. The emerging theme is that trait loss, or et al., 1977). Because stream capture events provide a time regression, can be the result of different evolutionary forces frame for the identification of discrete colonization events by and genetic and developmental pathways. Some genes appear A. mexicanus, divergence times of lineage splits between cave to be predisposed to be targets for evolutionary forces effecting populations and their most closely related surface populations morphological change (Protas et al., 2007), such as the pigmen- can be linked directly to the time of subterranean evolution. tation genes Oca2 and cinnabar. This may be due to a lack of Although it is not reasonable to assume that all of the traits deleterious pleiotropic connections and gene structure, leading evolve at the same time, based on hydrogeologic constraints, to similar morphologies evolving by independent mutations or fossil calibrations, and cytb sequences data, acquisition of the changes in expression patterns in the same genes. It remains troglomorphic form in A. mexicanus is estimated to be younger to be seen whether the patterns of evolutionary change high- than 2.2-5.2 Ma (Porter et al., 2007). It should be noted that

62 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Evolution in Karst Lineages

these times correlate to the time of subterranean colonization, Ptomaphagus beetles (Coleoptera, Leiodidae, Catopinae): An- not necessarily to the time of cave-associated trait acquisition. nales de Speleologie, v. 27, p. 399-404. However, estimating lineage splitting in A. mexicanus does Chippindale, P.T., Price, A.H., Wiens, J.J. and Hillis, D.M., place an upper limit on the time required for the acquisition of 2000, Phylogenetic relationships and systematic revision of the cave form, providing a working hypothesis for future stud- central Texas Hemidactyliine plethodontid salamanders: Her- ies of different species. petological Monographs, v. 14, p. 1-80.

CONCLUDING REMARKS Cooper, S.J.B., Bradbury, J.H., Saint, K.K., Leys, R., Austin, A.D. and Humphreys, W.F., 2007, Subterranean archipelago in The application of molecular tools to the study of subterra- the Australian arid zone: mitochondrial DNA phylogeography nean systems has led to new insights about the evolution of of amphipods from central Western Australia: Molecular Ecol- ogy, v. 16, p. 1533-1544. cave species and about the evolution of traits in general. It is clear from the studies that have been highlighted here that Cooper, S.J.B., Hinze, S., Leys, R., Watts, C.H.S. and Hum- to fully understand subterranean biodiversity a union of mo- phreys, W.F., 2002, Islands under the desert: molecular sys- lecular data on cryptic genetic diversity with classic taxonomy tematics and evolutionary origins of stygobitic water beetles is required; however, the methods needed to accomplish this (Coleoptera:Dytiscidae) from central Western Australia: Inver- are not yet well developed. Although much has been learned tebrate Systematics, v. 16, p. 589-598. about the evolution of trait loss from the A. mexicanus system, Espinasa, L., Yamamoto,Y. and Jeffery, W.R., 2005, Non- the interplay between trait loss and acquisition is also not well optical releasers for aggressive behavior in blind and blinded understood. As the evolution of the cave form is investigated Astyanax (Teleostei, Characidae): Behavioural Processes, v. in additional species, it will be fascinating to determine if the 70, p. 144-148. same mechanisms and forces as observed in A. mexicanus play a role in explaining the extreme convergence of cave adapta- Finston, T.L., Johnson, M.S., Humphreys, W.F., Eberhard, S.M. tion. Cave-adapted species remain a model system for studying and Halse, S.A., 2007, Cryptic speciation in two widespread basic evolutionary processes and the molecular-based studies subterranean amphipod genera reflects historical drainage pat- highlighted here illustrate that there is still much to be learned. terns in an ancient landscape: Molecular Ecology, v. 16, p. 355- 365.

REFERENCES Goricki, S. and Trontelj, P., 2006, Structure and evolution of the mitochondrial control region and flanking sequences in the Aden, E., 2005, Adaptation to darkness: In Encyclopedia of European cave salamander Proteus anguinus: Gene, v. 378, p. Caves, D.C. Culver and W.B. White [eds.], Amsterdam, El- 31-41. sevier Academic Press, p. 1-3. Hetrick, S.W. and Gooch, J.L., 1973, Genetic variation in popu- Avise, J.C., and Selander, R.K., 1972, Evolutionary genetics lations of the freshwater amphipod Gammarus minus (Say) in of cave-dwelling fishes of the genus Astyanax: Evolution, v. the central Appalachians. National Speleological Society Bul- 26, p. 1-19 letin, v. 35, p. 17-18.

Borowsky, R., 2008, Restoring sight in blind cavefish: Current Jeffery, W.R., 2001, Cavefish as a model system in evolution- Biology, v. 18, p. R23-R24. ary developmental biology: Developmental Biology, v. 231, p. 1-12. Buhay, J.E., and Crandall, K.A., 2005, Subterranean phylogeography of freshwater crayfishes shows extensive gene flow Jeffery, W.R., 2005, Adaptive evolution of eye degeneration in and surprisingly large population sizes: Molecular Ecology, 14, the Mexican blind cavefish: Journal of Heredity, v. 96, p. 185- p. 4259-4273. 196.

Buhay, J.E., Moni, G., Mann, N. and Crandall, K.A., 2007, Kane, T.C. and Richardson, R.C., 1985, Regressive evolution: Molecular taxonomy in the dark: Evolutionary history, phy- an historical perspective: The National Speleological Society logeography, and diversity of cave crayfish in the subgenus Bulletin, v. 47, p. 71-77. Aviticambarus, genus Cambarus: Molecular Phylogenetics and Evolution, v. 42, p. 435-448. Laing, C., Carmody, R.G. and Peck, S.B., 1976, Population genetics and evolutionary biology of the cave beetle Ptomapha- Carmody, G.R., Murphy, G. and Peck, S.B., 1972, Prelimi- gus hirtus: Evolution, v. 30, p. 484-497. nary studies on electrophoretic variation in cavernicolous

Frontiers of Karst Research 63

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Evolution in Karst Lineages

Lefébure, T., Douady, C.J., Gouy, M. and Gibert, J., 2006a, Proudlove, G., and Wood, P.J., 2003, The blind leading the Relationship between morphological taxonomy and molecular blind: cryptic subterranean species and DNA taxonomy: Trends divergence within Crustacea: proposal of a molecular threshold in Ecology and Evolution,, v. 18, p. 272-273. to help species delimitation: Molecular Phylogenetics and Evo- lution, v. 40, p. 435-447. Sadoglu, P., 1955, A mendelian gene for albinism in natural cave fish: Experientia, v. 13, p. 394-395. Lefébure, T., Douady, C.J., Gouy, M., Trontelj, P., Briolay, J. and Gibert, J., 2006b, Phylogeography of a subterranean am- Sadoglu, P., 1956, A preliminary report on the genetics of the phipod reveals cryptic diversity and dynamic evolution in ex- Mexican cave characins: Copeia, v. 1956, p. 113-114. treme environments: Molecular Ecology, v. 15, p. 1797-1806. Sadoglu, P., 1958, Mendelian inheritance in the hybrids be- Lefébure, T., Douady, C.J., Malard, F. and Gibert, J., 2007, tween the Mexican blind cave fishes and their overground Testing dispersal and cryptic diversity in a widely distributed ancestor. Verhandlungen Deutsche Zoologische Gesellschaft groundwater amphipod (Niphargus rhenorhodanensis): Molec- 1957, Supplement 21, p. 432-439. ular Phylogenetics and Evolution, v. 42, p. 676-686. Sbordoni, V., Sbordoni, M.C. and Matthaeis, E.D., 1979, Di- Leys, R., Cooper, S.J.B., Strecker, U. and Wilkens, H., 2005, vergenza genetica tra popolazioni e specie ipogee ed epigee di Regressive evolution of an eye pigment gene in independently Niphargus (Crustacea, Amphipoda): Lavori della Societa Itali- evolved eyeless subterranean diving beetles: Biology Letters, ana di Biogeografia, v.6, p. 329-351. v. 1, p. 496-499. Trontelj, P., 2007, The age of subterranean crayfish species. A Menuet, A., Alunni, A., Joly, J.-S., Jeffery, W.R. and Rétaux, comment on Buhay & Crandall (2005): subterranean phylo- S., 2007, Expanded expression of sonic hedgehog in Astyanx geography of freshwater crayfishes shows extensive gene flow cavefish: multiple consequences on forebrain development and and suprisingly large population sizes: Molecular Ecology, v. evolution: Development, v. 134, p. 845-855. 16, p. 2841-2843.

Mitchell, R.W., Russell, W.H. and Elliott, W.R., 1977, Mexican Verovnik, R., Sket, B. and Trontelj, P., 2004, Phylogeography eyeless characin fishes, genusAstyanax : environment, distribu- of subterranean and surface populations of water lice Asellus tion, and evolution: Texas Tech University Special Publications aquaticus (Crustacea: Isopoda): Molecular Ecology, v. 2004, of the Museum, v. 12, p. 1-89. p. 1519-1532.

Paquin, P., and Hedin, M., 2004, The power and perils of 'mo- Wiens, J.J., Chippindale, P.T. and Hillis, D.M., 2003, When are lecular taxonomy': a case study of eyeless and endangered phylogenetic analyses misled by convergence? A case study Cicurina (Araneae: Dictynidae) from Texas caves: Molecular in Texas cave salamanders: Systematic Biology, v. 52, p. 501- Ecology v. 13, p. 3239-3255. 514.

Parra-Olea, G., 2003, Phylogenetic relationship of the genus Wilcox, T.P., d.León, F.J.G., Hendrickson, D.A. and Hillis, Chiropterotriton (Caudata: Plethodontidae) based on 16S ribo- D.M., 2004, Convergence among cave catfishes: long-branch somal mtDNA: Canadian Journal of Zoology, v. 18, p. 2048- attraction and a Bayesian relative rates test: Molecular Phylo- 2060. genetics and Evolution, v. 31, p. 1101-1113.

Porter, M.L., Dittmar, K. and Pérez-Losada, M., 2007, How Wilkens, H., 1971, Genetic interpretation of regressive evolu- long does evolution of the troglomorphic form take? Estimat- tionary processes: studies on hybrid eyes of two Astyanax cave ing divergence times in Astyanax mexicanus: Acta Carsologica, populations (Characidae, Pisces): Evolution, v. 25, p. 530-544. v. 36, p. 173-182. Yamamoto, Y., Stock, D.W. and Jeffery, W.R., 2004, Hedgehog Protas, M., Conrad, M., Gross, J.B., Tabin, C. and Borowsky, signalling controls eye degeneration in blind cavefish: Nature, R., 2007, Regressive evolution in the Mexican Cave Tetra, v. 431, p. 844-847. Astyanax mexicanus: Current Biology, v. 17, p. 452-454. Zaksek, V., Sket, B. and Trontelj, P., 2007, Phylogeny of the Protas, M.E., Hersey, C., Kochanek, D., Zhou, Y., Wilkens, H., cave shrimp Troglocaris: evidence of a young connection be- Jeffery, W.R., Zon, L.I., Borowsky, R. and Tabin, C.J., 2006, tween Balkans and Caucasus: Molecular Phylogenetics and Genetic analysis of cavefish reveals molecular convergence in Evolution, v. 42, p. 223-235. the evolution of albinism: Nature Genetics, v. 38, p. 107-111.

64 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Karst Resources and Other Applied Issues

Karst Resources and Other Applied Issues

Dorothy J. Vesper Department of Geology West Virginia University Morgantown, WV 26506

INtRODUCTION agement for sustainable-resource use requires that we better understand the fundamental processes in karst systems so that pplied issues in karst span a wide range of topics, in- we can better translate between them and better predict likely cluding land-use planning and impacts, availability of impacts. water resources, flooding, subsidence, building stabil- A The applied issues are divided into three categories by type: (1) ity, and degradation of water quality. The importance of these issues cannot be overstated given the large number of people water quality, (2) water quantity, and (3) geotechnical issues. that reside in karst regions and ecosystems that rely on karst The relevance and implication related to land use at defense environments and water to sustain life. Furthermore, we need installations are also discussed. to conserve karst systems for use in studying large scientific questions and fundamental processes. WATER QUALITY

Most of the work on applied issues has been focused on case Water quality can be degraded by the transport of suspended studies and solving specific problems: for example, remediat- sediments or the introduction and transport of chemical con- ing sinkholes and subsidence for engineering purposes, de- taminants and biological agents. These can be input into the creasing flood risks, and tracking aquifer contaminants. The system either quickly via sinking streams, sinkholes, and en- site-specific nature of an applied approach provides important hanced vertical pathways or slowly via a more dispersed route local solutions, but it does not always address the fundamental (Figure 1). Spatial distribution ranges from individual point- scientific questions or mechanisms that can allow the answer to sources such as spills and underground tanks to non-point be generalized to other systems. sources such as widespread agricultural spraying. Once in the aquifer, contaminant fate and transport is dependent on the Many of the issues related to land and water resources overlap type of contaminant and its chemical properties, as well as the with research being conducted in non-karst areas. However, chemical and physical properties of the aquifer. Transport of karst systems have some unique characteristics that require di- contaminants can range from rapid movement (velocity of me- rected attention and research: ters per second) through the system and discharge at springs, to trapping mechanisms that result in long-term storage (velocity • Close connections between surface and subsurface pro- of less than a meter per year) with little or only very gradual cesses render karst systems highly vulnerable to impacts release. Continuous monitoring can help capture these highly from surface activities. variable release mechanisms. New model development needs • Spatial heterogeneity hinders our ability to easily monitor to incorporate functions to better interpret these processes and and assess water quality and quantity. the varied contaminants that may be present in karst. • The rates of physical processes are highly variable. • Subsidence may occur at an almost imperceptible rate or Sediment Transport be nearly instantaneous. Contaminants may be rapidly flushed through the system or trapped indefinitely. Water Numerous recent studies have focused on the presence and flow at springs may be consistent through time or change transport of sediments in karst systems. How and when sedi- rapidly in response to storm events. ments move are closely linked to water quality from both a turbidity and contaminant transport standpoint. Many types of Although karst systems have some unique characteristics, these contaminants associate with solids and may be either stored or topics provide insight to other heterogeneous sites such as frac- transported along with the sediments. The current direction of tured rock and urban systems with pipe infrastructure. Man- the research is looking at storage of sediments and larger materials within the system, suspended sediment flux during storms,

Frontiers of Karst Research 65

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Karst Resources and Other Applied Issues

Figure 1. Karst trapping mechanisms and transport styles for light and dense non-aqueous phase liquids (LNAPL and DNAPL, respectively).

the nature of the transported sediments, and mathematical strat- Autocorrelation and spectral analysis are being used to deter- egies for comparing sediment transport to other parameters. mine the relationship between turbidity and discharge (Boucha- ou et al., 2002; Padilla and Pulido-Bosch, 1995). Time series Transport is typically linked to turbulent flow, and thus to storm analysis can help identify different time components in the data events. Both particle size (Atteia and Kozel, 1997) and min- and residence time of the input-pulse (a storm) through the sys- eralogy (Mahler and Lynch, 1999) can change over the course tem. of a storm. Studies have identified sediments transported from the surface during storms (Ryan and Meiman, 1996) and due to Transport of particles is poorly explained through the use of re-suspension of sediments already in the system (Herman et traditional dissolved tracers. Therefore, the design and testing al., 2008; Marshall et al., 1998; Pronk et al., 2006). of solid-phase tracers transported in suspension is an active area of research. Success has been demonstrated for lanthanide- The relationship between discharge and turbidity varies depen- labeled clay (Mahler et al., 1998; Ting, 2005), DNA-labeled dent on antecedent conditions (Toran et al., 2006). Threshold clay (Mahler et al., 1998), bacteria (Auckenthaler et al., 2002; discharge values may exist for sediment transport and re-mo- Dussart-Baptista et al., 2003), europium-tagged bacteria (Ting, bilization of larger materials already within the system (White, 2005), and microspheres (Auckenthaler et al., 2002). The 1988). A threshold-crossing event was recently observed at lanthanide-labeled clay tracer (Mahler et al., 1998) was trans- Pennsylvania spring, following three major hurricanes, when ported faster than the co-injected dissolved solute (NaCl for the sediment discharged at a spring increased by nearly an surface water, rhodamine for karst ground water). Other re- order-of-magnitude above previous records (Herman et al., searchers have observed similar results suggesting this may be 2008). due to matrix exclusion of particles injected into the soil zone

66 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Karst Resources and Other Applied Issues

(Auckenthaler et al., 2002), although this does not explain the (Vesper and White, 2003, 2004). The neutral to alkaline pH surface water results for the lanthanide-labeled clays. These of karst settings renders most metals insoluble via either pre- studies argue strongly for the need for coupled tracers to clarify cipitation or sorption. Elevated metal concentrations have been different modes of transport. found in urban spring sediments (Gutiérrez et al., 2004) and springs downgradient from metal contaminant sources (Vesper Inorganic Contaminants and White, 2004).

Inorganic contaminants are among the most intensively-studied Microbes and Pathogens of the contaminant groups. Some inorganic compounds, such as nitrates and chlorides, are highly water soluble. Others, such Microbes are readily transported through karst aquifers. They as heavy metals, are more typically associated with solids ei- have been shown to enter the aquifer from surface sources ther through precipitation or sorption. (Marshall et al., 1998; Ryan and Meiman, 1996) and from pumping-induced surface-water infiltration (Mahler et al., Nitrate and nutrients have widespread sources in agriculture, 2000). Re-suspension of fecal coliform already in the system industry, and wastewater. Nitrate has been shown to vary in has also been suggested (Davis et al., 2005; Marshall et al., concentration over several time scales. Storm-based studies 1998). Inoculated chambers placed in karst springs and streams have shown that nitrate can either increase or decrease during showed that Excherichia coli can survive up to four months af- storms, depending on the source type. Nitrate associated with ter introduction into the system (Davis et al., 2005). surface contamination, such as feed lots, increases at springs during storm events as the overland flow injects the contami- Although coliform concentrations have been shown to increase nant into the aquifer. If the nitrate is stored within the sys- during storms (Ryan and Meiman, 1996), Crytosporidium par- tem and continuously released, its concentration can decrease vum oocysts have been found in spring water during baseflow during storms. Nitrate can also change seasonally with land conditions (Kuczynska et al., 2003). Several studies have indi- application of fertilizers (Panno and Kelly, 2004). Isotopic cated that although microbial concentrations in springs are gen- studies of the δ 15N and δ 18O in the nitrate ion link contami- erally high during storms when the turbidity increases, bacteria nation to fertilizers that have undergone some denitrification can also be present during baseflow or low-turbidity conditions (Panno et al., 2001). Strategies to predict nitrate concentrations (Dussart-Baptista et al., 2003; Pronk et al., 2006). A compari- in springs have included multi-regression models (Peterson et son of sessile (attached) and planktonic (non-attached) bacteria al., 2000), autocorrelation analysis (Jones and Smart, 2005), transport from a sinkhole to a spring found that the planktonic and tributary-mixing models (Perrin et al., 2006). All of the bacteria were more closely correlated to turbidity than the ses- approaches produced results that were site-specific, suggesting sile bacteria (Dussart-Baptista et al., 2003). that the flow-nitrate concentration relationship is not readily generalized. Anthropogenic Organic Compounds

Another effluent and agriculturally-related contaminant receiv- Organic compounds have a wide range of chemical charac- ing recent attention is phosphorous which has been studied teristics and may preferentially partition into numerous phys- in Irish springs (Kilroy and Coxon, 2005). Phosphorous was iochemical compartments: gases, dissolved in water, incorpo- found to increase during storm events and travel primarily in rated into solids, sorbed to organic or inorganic solids, in biota, the particulate form – indicating its transport might be closely or as free-phase product. This in turn controls their transport associated with suspended sediment movement. and storage. The distribution can be estimated using the chemical properties of the compound (vapor pressure, solubility, ac- Chloride and sodium are highly soluble ions often attributed id-base dissociation) and partitioning coefficients (Henry’s law to urban and road salt sources. These concentrations are often constant defines the relative concentrations between air and highest in the springtime (Panno et al., 2001). Retardation of water, octanol-water portioning coefficients as proxy for mass sodium relative to chloride, in a road-salt contaminated frac- transfer between organic phases and water, various solid-water tured aquifer, has been attributed to temporary sequestration partitioning coefficients). via ion exchange (Werner and diPretoro, 2006). Knowledge of the environmental setting is critical. For ex- Heavy metals and metalloids typically associate with solids ample, solubility is a function of temperature, solution ionic and are transported in conjunction with suspended sediments strength, and the presence of co-solutes; sorption to solids is

Frontiers of Karst Research 67

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Karst Resources and Other Applied Issues

closely tied to the fraction of organic carbon present. In addi- of chloroform and styrene from a train derailment to Wilson tion to the transfer of contaminants between physiochemical Spring, Tennessee (Crawford and Ulmer, 1994). Wells installed or system compartments, transformation reactions can occur. in the aquifer identified DNAPL chloroform and LNAPL sty- If contaminants are destroyed (degraded), product compounds rene near the spill. The DNAPL migrated down dip along a can be created which are also hazardous. confining layer rather than in the direction of ground water flow. Wilson Spring discharged low levels of both compounds Studies focused on the presence and transport of organic com- and, after large storms, styrene product. A more recent study pounds in karst settings are increasing. In additional to the of springs contaminated by VOCs (Williams and Farmer, 2003; temporal variation found for all water quality parameters, low- Williams et al., 2006) found that Wilson Spring continues to solubility organic compounds can be present as non-aqueous discharge chloroform and styrene. Chloroform concentrations phase liquids (NAPLs). increase during storms, with the greatest change after first-flush storms. At Cascade Spring, Tennessee, contaminated with cis- Pesticide concentrations in springs vary seasonally with land 1,2-dichloroethene (Williams and Farmer, 2003; Williams et application (Panno and Kelly, 2004; Pasquarell and Boyer, al., 2006), the VOC concentration was observed to decrease 1996). The presence of atrazine on suspended sediments in- during storms. dicates that pesticides may be stored in aquifer sediments and transported periodically (Panno and Kelly, 2004). The storage Overarching Water Quality Issues Unique to and breakdown of atrazine in the soil zone was tracked sea- Karst Systems sonally by the ratio of atrazine to its degradation–product des- ethylatrazine (DES) (Pasquarell and Boyer, 1996). PCBs were Several issues exist for all types of water-quality parameters studied at a spring in central Indiana and found to increase in and are unique to karst systems: concentration during storms (Krothe and Fei, 1999). Natural estrogen, a tracer for waste products, has been measured in the • Source location and mode of introduction often controls form of 17 β-estrodial in eight Missouri springs (Wicks et al., patterns of contamination. It is possible for contaminant 2004) and five Arkansas springs (Peterson et al., 2000). Peri- injection via sinkholes or exposed epikarst during storm odic samples from Meramec Spring in Missouri indicate that events. This process can rapidly introduce mobile sedi- the concentration increases with discharge. ments, soluble contaminants, particulate and solid-associated contaminants, and microbial pathogens. Flow systems NAPL contamination can exist either as residual (held to grains of this type can result in highly variable contaminant con- via surface tension) or pooled (Fig. 1). Both forms act as long- centrations at downgradient locations. Infiltration through term sources, slowly releasing contaminants into water. Light a thick soil cover can also lead to long-term sources that NAPLs (LNAPL) are less dense than water and float on the are mobilized episodically. top of the water table. They may be trapped or decanted in a • Temporal variability is a common characteristic of con- karst aquifer due to the physical geometry of the flow system taminant transport in karst. Whereas some contaminant (Ewers et al., 1992) (Fig. 1). At Fort Campbell Army Airfield, concentrations may increase during storms, others have a 16-foot thick layer of jet fuel was documented, although no been shown to decrease. The direction of change may de- free product was present in Quarles Spring – the traced output pend on the contaminant properties and where it is stored from the airfield (Ewers et al., 1992). Later studies of Quarles within in the system. Well-defined contaminant plumes Spring indicated the presence of a “well washed jet fuel” as- are uncommon; episodic transport is the norm. sociated with the spring sediments with only low concentra- • Karst systems have NAPL trapping mechanisms not typi- tions of volatile organic compounds (VOCs) in the water (Ves- cally found in other types of aquifers. These can cause per and White, 2006). Ewers and others also investigated the storage and retardation of contaminants and hinder reme- transport of LNAPL at a site in Richmond, Kentucky. A leaky diation attempts. underground storage tank was traced to two springs using dyes; however, only one of the springs contained detectable VOCs WATER QUANTITY (Ewers et al., 1992). Knowing the sustainable yield for a karst aquifer, and how Dense NAPL (DNAPL) sinks during transport, resulting in a climate and land-use activities impact that yield, is essential flowpath that may not match the overall direction of ground information for water-resource management. Unfortunately, water flow (Fig. 1). Crawford and Ulmer tracked a 1990 spill many of the techniques used to determine storage and sustain-

68 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Karst Resources and Other Applied Issues

able yield in porous media aquifers (e.g., pump tests) are not and have created millions of dollars in losses and a number of directly applicable to karst aquifers or other highly heteroge- fatalities. Sinkhole collapse has resulted in the loss of structural neous systems. Recent attempts at evaluating aquifer yield support of bridge foundations, roadways, railways, and build- have employed water balances, spring flow and hydrograph ings and has occurred in many states. Ten Multidisciplinary data, and age-dating to determine the residence time of water Conferences on Sinkholes & Engineering & Environmental underground. Impacts of Karst have occurred since 1984; the proceedings contain numerous excellent case studies. Failure to take into Detailed water balances have been determined in some places. consideration karst processes in the location and construction A water balance was determined for the Trnovsko-Banjška of large reservoirs commonly result in either extreme cost over- Planota Plateau in Slovenia (Trisic, 1997). Spring flow data runs in projects and/or the failure of the structures to perform and hydrograph separation has been used to estimate baseflow as designed. and storage in karst aquifers. However, spring discharge may be limited due to conduit size, connections between basins dur- Like the other applied issues, most of the research in geotechni- ing high flow, delivery to intermittent springs, and dynamic cal aspects of karst has been focused on local problem mitiga- epikarstic storage (Bonacci, 2001). Land use and engineer- tion and engineering solutions. Risk analysis and the mapping ing change can cause significant decrease in sustainable wa- of hazard areas have been active topics. ter yield. Storage losses have been witnessed in China due to deforestation in the upland areas of the aquifers in the Stone Active research topics in applied geotechnical issues include Forest (Huntoon, 1992). Without the flow-buffering capacity techniques for identifying underground karst features, mecha- of storage, water resources have become more variable, with nisms of subsidence and collapse, and stability of dams: increases in both droughts and floods. • Various geophysical techniques have been investigated for Long-term spring monitoring pre- and post-dam building in their use in locating sinkholes and underground structures. Herzegovina found that the minimum spring discharge did not Some success has been obtained through use of resitivity decrease but the annual average output did owing to rerouting and microgravity (McGrath et al., 2002; Roth et al., 2002). of recharge water into engineered structures (Milanovic, 2002). 2-D electrical resitivity tomography has been recently used Other studies have indicated that urbanization can increase to delineate vadose flowpaths in sinkholes (Schwartz and aquifer recharge owing to the increased infiltration from utility Schreiber, 2006). Smoke tests have been combined with infrastructure (Lerner, 2002). geophysical techniques to map the vadose karst (Nyquist et al., 2005) and determine if cavities are open or closed Long-term yield has also been addressed through use of atmo- (Meighan et al., 2006). Improvements in data interpreta- spheric tracers used to identify the residence time of water un- tion are also being advanced (Nyquist et al., 2007; Nyquist derground or the mixing of old and young waters. Low tritium and Roth, 2005). concentrations during baseflow in Slovenian springs indicate • The mechanics of sinkhole collapse have been modeled the discharge of storage water greater than 50 years old (Bi- and illustrate the importance of recharge seepage pressure ondic et al., 2006). A suite of tracers, including chlorofluoro- and over-pumping (Keqiang et al., 2004). Physical mod- carbons, have been used to date the shallow water component els have been used to predict failure in cemented sands in the Floridan aquifer (Plummer et al., 1998; Plummer et al., based on sand cohesion and thickness; and cavity geom- 1998). etry (Goodings and Abdulla, 2002). • Leakage of water from dams built in karst areas is a well The storage capacity and sustainable yield of a karst aquifer is established problem (Milanovic, 2002). Research on this not readily determined (Bakalowicz, 2005). When karst wa- topic utilized coupled flow-geochemical models indicates ter supplies are mined or over-exploited, impacts can include that the increase in hydraulic gradient created by the dam salinization, intensification of karst processes, and catastrophic drives the dissolution rates (Dreybrodt et al., 2002; Ro- subsidence (Pulido-Bosch, 1999). manov et al., 2003).

GEOTECHNICAL PROBLEMS U.S. MILITARY INSTALLATIONS ON KARST Naturally occurring processes such as sinkhole collapse or sinkhole flooding can be accelerated or induced by human activities Military bases typically have similar land uses as are found in

Frontiers of Karst Research 69

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Karst Resources and Other Applied Issues

metropolitan areas, such as residential areas, sewers, roadways, Most military installations that have well-developed karst have and airports. The major difference is military-specific use and already taken their karst setting into account for water resource management of ammunition and ordnance. The Department and environmental planning. Some installations in karst have of Defense (DoD) has established the Defense Environmental used, or been part of, a basin-scale approach for water resourc- Restoration Program (DERP) to address these issues. Within es or environmental impact. Examples of these studies include that structure, the Installation Restoration Program (IRP) is fo- extensive dye tracing to frame conceptual models and delineate cused on health and safety impacts related to contamination. ground-water basins at numerous karst facilities: Fort Knox The DoD reports there are 4,200 properties in the IRP (U.S. Military Reservation, Kentucky (Connair and Murray, 2002), Department of Defense, 2007). Fort Campbell, Kentucky/Tennessee (Arthur D. Little Inc., 1997), and Fort Leonard Wood Military Reservation, Missouri A large number of military bases in the contiguous U.S. (more (Imes et al., 1996; Kleeschulte and Imes, 1997). Other bases than 70) are located in karst regions (Fig. 2). All of the applied have incorporated a karst-approach to their regulatory investi- topics above – water quality, water quantity, geotechnical is- gation. For example, the Letterkenny Army Depot in southcen- sues – are relevant to these sites. Many of the bases use karst tral Pennsylvania is listed on EPA’s National Priorities List and water as their primary drinking water supply. At Fort Campbell is one of the 1995 BRAC sites (U.S. Environmental Protection Army Base, 40,000 customers use water from Boiling Spring Agency, 2007). Chlorinated VOCs have been detected in some (Fort Campbell, 2007). For bases being closed under the Base springs on the facility. Karst springs were also found to dis- Realignment and Closure (BRAC) program, planning for land charge nitrates and RDX associated with ammunition burning re-use requires knowledge of contaminants present, environ- at the Crane Naval Airfield Warfare Center in Indiana (DiG- mental liabilities, and vulnerable areas. nazio et al., 1998; Krothe, 2003). San Antonio, on the edge of the karstic Edwards Aquifer, is home to multiple military facilities. One of those facilities, Camp Bullis, has conducted

Figure 2. Location of karst regions (shaded) and military bases (dots) in the contiguous U.S. Locations of examples indicated. Modified from the National Karst Map and the National Park Service (National Cave and Karst Research Institute and U.S. Geological Survey, 2007; U.S. National Park Service, 2007)

70 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Karst Resources and Other Applied Issues

integrated assessments of water resources, hydrogeology, caves DiGnazio, F.J., Krothe, N.C., Baedke, S.J., and Spalding, R.F., and endangered species (Veni, 2004). 1998, δ 15N of nitrate derived from explosive sources in a karst aquifer beneath the Ammunition Burning Ground, Crane Naval There are also U.S. military sites outside the country in which Surface Warfare Center, Indiana USA: Journal of Hydrology, v. 206, p. 164-175. karst systems have been impacted. The Anderson Air Force Base in northern Guam is listed on the EPA National Prior- Dreybrodt, W., Romanov, D., and Gabrovsek, F., 2002, Karsti- ity List. Ground water is contaminated with lead, chlorinated fication below dam sites: a model of increasing leakage from solvents and toluene. The underlying North Guam Lens is the reservoirs: Environmental Geology, v. 42, p. 518-524. primary drinking water source for the area (U.S. Agency for Toxic Substances and Disease Registry, 2002). Dussart-Baptista, L., Massei, N., Dupoint, J.-P., and Jouenne, T., 2003, Transfer of bacteria-contaminated particles in a karst REFERENCES aquifer: evolution of contaminated materials from a sinkhole to a spring: Journal of Hydrology, v. 284, p. 285-295. Arthur D. Little Inc., 1997, The 1997 Hydrogeology Update Ewers, R.O., Duda, A.J., Estes, E.K., Idstein, P.J., and Johnson, Report, Fort Campbell Kentucky, Submitted to the Hazardous K.M., 1992, The transmission of light hydrocarbon contami- Waste Remedial Actions Program, U.S. Department of Energy, nants in limestone (karst) aquifers: Third Conference on Hy- p. 641. drogeology, Ecology, Monitoring and Management of Ground Waters in Karst Terranes: Nashville, TN, Water Well Journal Atteia, O., and Kozel, R., 1997, Particle size distributions in Publishing Company, p. 287-304. water from a karstic aquifer: from particles to colloids: Journal of Hydrology, v. 201, p. 102-119. Fort Campbell, 2007, Fort Campbell Environmental Program, Water Programs. Auckenthaler, A., Raso, G., and Huggenberger, P., 2002, Par- ticle transport in a karst aquifer: natural and artificial tracer Goodings, D.J., and Abdulla, W.A., 2002, Stability charts for experiments with bacteria, bacteriophages and microspheres: predicting sinkholes in weakly cemented sand over karst lime- Water Science and Technology, v. 46, p. 131-138. stone: Engineering Geology, v. 65, p. 179-184. Bakalowicz, M., 2005, Karst groundwater: a challenge for new Gutiérrez, M., Neill, H., and Grand, R.V., 2004, Metal in sedi- resources: Hydrogeology Journal, v. 13, p. 148-160. ments of springs and cave sediments as environmental indicators in karst areas: Environmental Geology, v. 46, p. 1079- Biondic, B., Biondic, R., and Kapelj, S., 2006, Karst ground- 1085. water protection in the Kupa River catchment area and sustainable development: Environmental Geology, v. 49, p. 828-839. Herman, E.K., Toran, L., and White, W.B., 2008, Threshold events in spring discharge: evidence from sediment and con- Bonacci, O., 2001, Analysis of the maximum discharge of karst tinuous water level measurement: Journal of Hydrology, v. 351, springs: Hydrogeology Journal, v. 9, p. 328-338. p. 98-106. Bouchaou, L., Mangin, A., and Chauve, P., 2002, Turbidity Huntoon, P.W., 1992, Hydrogeologic characteristics and de- mechanism of water from a karstic spring: example of the Ain forestation of the stone forest karst aquifers of south China: Asserdoune spring (Beni Mellal Atlas, Morocco): Journal of Ground Water, v. 30, p. 167-175. Hydrology, v. 265, p. 34-42. Imes, J.L., Schumacher, J.G., and Kleeschulte, M.J., 1996, Connair, D.P., and Murray, B.S., 2002, Karst groundwater ba- Geohydrologic and water-quality assessment of the Fort Leon- sin delineation, Fort Knox, Kentucky: Engineering Geology, v. ard Wood Military Reservation, Missouri, 1994-95, U.S. Geo- 65, p. 125-131. logical Survey, p. 134. Crawford, N.C., and Ulmer, C.S., 1994, Hydrogeologic investi- Jones, A.L., and Smart, P.L., 2005, Spatial and temporal chang- gations of contaminant movement in karst aquifers in the vicin- es in the structure of groundwater nitrate concentration time ity of a train derailment near Lewisburg, Tennessee: Environ- series (1935-1999) as demonstrated by autoregressive model- mental Geology, v. 21, p. 41-52. ling: Journal of Hydrology, v. 310, p. 201-215.

Davis, R.K., Hamilton, S., and Brahana, J.V., 2005, Escheri- Keqiang, H., Bin, W., and Dunyum, Z., 2004, Mechanism and chia coli survival in mantled karst springs and streams, north- mechanical model of karst collapse in an over-pumping area: west Arkansas Ozarks, USA: Journal of the American Water Environmental Geology, v. 46, p. 1102-1107. Resources Association, v. 41, p. 1275-1287.

Frontiers of Karst Research 71

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Karst Resources and Other Applied Issues

Kilroy, G., and Coxon, C., 2005, Temporal variability of phos- tools for water resource investigations in karst terrain: Environ- phorus fractions in Irish karst springs: Environmental Geology, mental Geology, v. 42, p. 552-557. v. 47, p. 421-430. Meighan, H., Nyguist, J.E., and Roth, M.J.S., 2006, Mise-á- Kleeschulte, M.J., and Imes, J.L., 1997, Regional ground- masse, resistivity tomography and smoke tests combined to water flow directions and spring recharge in and near the Fort map karst, Easton, PA: Geological Society of America Ab- Leonard Wood Military Reservation, Missouri: Rolla MO, U.S. stracts with Programs, v. 38, p. 525. Geological Survey, Fact Sheet FS-101-97, p. 4. Milanovic, P., 2002, The environmental impacts of human ac- Krothe, N.C., 2003, Groundwater flow and contaminant trans- tivities and engineering construction in karst regions: Episodes, port through epikarst in two karst drainage systems, USA: v. 25, p. 13-21. RMZ - Materials and Geoenvironment, v. 50, p. 177-180. National Cave and Karst Research Institute, and U.S. Geologi- Krothe, N.C., and Fei, Y., 1999, Polychlorinated byphenyl cal Survey, 2007, The National Karst Map. (PCB) contamination of a karst aquifer in an urban environment, Central Indiana, USA, in Chilton, J., ed., Groundwater Nyquist, J.E., Peake, J., and Roth, M.J.S., 2007, Comparison of in the Urban Environment: Rotterdam, A.A. Balkema, p. 171- an optimized resistivity array with dipole-dipole soundings in 179. karst terrain: Geophysics, v. 72, p. F139-F144.

Kuczynska, E., Boyer, D.G., and Shelton, D.R., 2003, Com- Nyquist, J.E., and Roth, M.J.S., 2005, Improved 3D pole-dipole parison of immunofluorescence assay and immunomagnetic resistivity surveys using radial measurement pairs: Geophysi- electrochemiluminescence in detection of Cryptosporidium cal Research Letters, v. 32, p. L21504. parvum oocysts in karst water samples: Journal of Microbio- logical Methods, v. 53, p. 17-26. Nyquist, J.E., Roth, M.J.S., Henning, S., Manney, R., and Peake, J., 2005, Smoke without mirrors: a new tool for the geo- Lerner, D.N., 2002, Identifying and quantifying urban recharge: physical characterization of shallow karst cavities, Proceedings a review: Hydrogeology Journal, v. 10, p. 143-152. of the SAGEEP -05 Symposium for the Application of Geo- physics to Environmental and Engineering Problems, p. 337- Mahler, B.J., Bennett, P.C., and Zimmerman, M., 1998, Lan- 343 published on CDRom. thanide-labeled clay: a new method for tracing sediment transport in karst: Ground Water, v. 36, p. 835-843. Padilla, A., and Pulido-Bosch, A., 1995, Study of hydrographs of karstic aquifers by means of correlation and cross-spectral Mahler, B.J., and Lynch, F.L., 1999, Muddy waters: temporal analysis: Journal of Hydrology, v. 168, p. 73-89. variation in sediment discharging from a karst spring: Journal of Hydrology, v. 214, p. 165-178. Panno, S.V., Hackley, K.C., Hwang, H.H., and Kelly, W.R., 2001, Determination of the sources of nitrate contamination in Mahler, B.J., Personné, J.C., Lods, G.F., and Drogue, C., 2000, karst springs using isotopic and chemical indicators: Chemical Transport of free and particulate-associated bacteria in karst: Geology, v. 179, p. 113-128. Journal of Hydrology, v. 238, p. 179-193. Panno, S.V., and Kelly, W.R., 2004, Nitrate and herbicide load- Mahler, B.J., Winkler, M., Bennett, P.C., and Hillis, D.M., ing in two groundwater basins in Illinois sinkhole plain: Jour- 1998, DNA-labeled clay: a sensitive new method for tracing nal of Hydrology, v. 290, p. 229-242. particle transport: Geology, v. 26, p. 831-834. Pasquarell, G.C., and Boyer, D.G., 1996, Herbicides in karst Marshall, D., Brahana, J.V., and Davis, R.K., 1998, Resuspen- groundwater in southeast West Virginia: Journal Environmen- sion of viable sediment-bound enteric pathogens in shallow tal Quality, v. 25, p. 755-765. karst aquifers, in Brahana, J.V., Eckstein, Y., Ongley, L.K., Sch- Perrin, J., Jeannin, P.Y., and Cornaton, F., 2006, The role of neider, R., and Moore, J.E. eds., Gambling with Groundwater - tributary mixing in chemical variations at a karst spring, Milan- Physical, Chemical, and Biological Aspects of Aquifer-Stream dre, Switzerland: Journal of Hydrology, v. 332, p. 158-173. Relations: Proceedings Volume of the International Association of Hydrogeologists Congress XXVIII and the Annual Meeting Peterson, E.W., Davis, R.K., and Brahana, J.V., 2000, The use of the American Institute of Hydrology, Las Vegas: Smyrna of regression analysis to predict nitrate-nitrogen concentrations GA, The American Institute of Hydrology, p. 179-186. in springs of northwest Arkansas, in Sasowsky, I.D., and Wicks, McGrath, R.J., Styles, P., Thomas, E., and Neale, S., 2002, In- C.M., eds., Groundwater Flow and Contaminant Transport in tegrated high-resolution geophysical investigations as potential Carbonate Aquifers: Rotterdam, Netherlands, A.A. Balkema, p. 43-64.

72 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Karst Resources and Other Applied Issues

Peterson, E.W., Davis, R.K., and Orndorff, H.A., 2000, Trisic, N., 1997, Investigations of the water balance: Acta Car- 17β-Estradiol as an indicator of animal waste contamination sologica, v. 28, p. 123-141. in mantled karst aquifers: Journal of Environmental Quality, v. 29, p. 826-834. U.S. Agency for Toxic Substances and Disease Registry, 2002, Public Health Assessment, Andersen Air Force Base, Yigo, Plummer, L.N., Busenberg, E., Drenkard, S., Schlosser, P., Guam, ATSDR. Ekwurzel, B., Weppernig, R., McConnell, J.B., and Michel, R.L., 1998, Flow of river water into a Karstic limestone aqui- U.S. Department of Defense, 2007, Office of the Deputy Un- fer. 2. Dating the young fraction in groundwater mixtures in the dersecretary of Defense, Environmental Management Office, Upper Floridan aquifer near Valdosta, Georgia: Applied Geo- Defense Environmental Restoration Program, Volume 2007. chemistry, v. 13, p. 1017-1043. U.S. Environmental Protection Agency, 2007, Letterkenny Plummer, L.N., Busenberg, E., McConnell, J.B., Drenkard, S., Army Depot, Property Disposal Office. Schlosser, P., and Michel, R.L., 1998, Flow of river water into a Karstic limestone aquifer. 1. Tracing the young water fraction U.S. National Park Service, 2007, Military Bases in the Con- in groundwater mixtures in the Upper Floridan Aquifer near tiguous United States, Volume 2007. Valdosta, Georgia: Applied Geochemistry, v. 13, p. 995-1015. Veni, G., 2004, Multidisciplinary karst research on a military Pronk, M., Goldscheider, N., and Zopfi, J., 2006, Dynamics and reservation: Camp Bullis, Texas: Geological Society of Ameri- interaction of organic carbon, turbidity and bacteria in a karst ca Abstracts with Programs, v. 36, p. 191. aquifer system: Hydrogeology Journal, v. 14, p. 473-484. Vesper, D.J., and White, W.B., 2003, Metal transport to karst Pulido-Bosch, A., 1999, Chapter 7. Karst water exploitation, springs during storm flow: An example from Fort Campbell, in Drew, D., and Hötzl, H., eds., Karst Hydrogeology and Hu- Kentucky/Tennessee, U.S.A.: Journal of Hydrology, v. 276, p. man Activities; Impacts, Consequences and Implications: Rot- 20-36. terdam, A.A. Balkema, p. 225-256. Vesper, D.J., and White, W.B., 2004, Spring and conduit sedi- Romanov, D., Gabrovsek, F., and Dreybrodt, W., 2003, Dam ments as storage reservoirs for heavy metals in karst aquifers: sites in soluble rocks: a model of increasing leakage by disso- Environmental Geology, v. 45, p. 481-493. lution widening of fractures beneath a dam: Engineering Geol- ogy, v. 70, p. 17-35. Vesper, D.J., and White, W.B., 2006, Comparative storm response of contaminants in a carbonate aquifer, Fort Campbell, Roth, M.J.S., Mackey, J.R., Mackey, C., and Nyquist, J.E., Kentucky-Tennessee, in Harmon, R.S., and Wicks, C., eds., 2002, A case study of the reliability of multielectrode earth re- Perspectives on karst geomorphology, hydrology, and geo- sistivity testing for geotechnical investigations in karst terrains: chemistry, Geological Society of America Special Paper 404, Engineering Geology, v. 65, p. 225-232. p. 267-274.

Ryan, M., and Meiman, J., 1996, An examination of short- Werner, E., and diPretoro, R.S., 2006, Rise and fall of road salt term variations in water quality in a karst spring in Kentucky: contamination of water-supply springs: Environmental Geol- Ground Water, v. 34, p. 23-30. ogy, v. 51, p. 537-543.

Schwartz, B.F., and Schreiber, M., 2006, Integrating differen- White, W.B., 1988, Geomorphology and Hydrology of Karst tial electrical resistivity tomography and time domain reflec- Terrains: New York, Oxford University Press, 464 p. tometry as a tool for modeling soil moisture and infiltration in sinkholes: Geological Society of America Abstracts with Pro- Wicks, C.M., Kelley, C., and Peterson, E.W., 2004, Estrogen in grams, v. 38, p. 194. a karstic aquifer: Ground Water, v. 42, p. 384-389.

Ting, T.E., 2005, Assessing bacterial transport, storage and vi- Williams, S.D., and Farmer, J.J., 2003, Volatile organic com- ability in mantled Karst of northwest Arkansas using clay and pound data from three karst springs in Middle Tennessee, Feb- Escherichia coli labeled with lanthanide series metals [Ph.D. ruary 2000 to May 2001: Nashville, TN, U.S. Geological Sur- thesis]: Fayetteville, AK, University of Arkansas. vey, Open-File Report 03-355, p. 69.

Toran, L., Tancredi, J.H., Herman, E.K., and White, W.B., Williams, S.D., Wolfe, W.J., and Farmer, J.J., 2006, Sampling 2006, Conductivity and sediment variation during storms as strategies for volatile organic compounds at three karst springs evidence of pathways to karst springs, in Harmon, R.S., and in Tennessee: Ground Water Monitoring & Remediation, v. 26, Wicks, C., eds., Perspectives on karst geomorphology, hydrol- p. 53-62. ogy, and geochemistry, Geological Society of America Special Paper 404, p. 169-176.

Frontiers of Karst Research 73

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13

Part 3

FINDINGS AND RECOMMENDATIONS OF THE FOCUS GROUPS FINDINGS AND RECOMMENDATIONS OF THE FOCUS GROUPS

The discussions of the Focus Groups were recorded as notes Groups were encouraged to present their findings in any way and on large paste-up sheets on the walls of the meeting room. that seemed appropriate. Specific topics were identified and these were discussed in several iterations. At the end, the Focus Group leaders drew the The names of the Focus Group participants are listed with the main points together as the written reports in the sections that reports. Their affiliations, addresses, and other contact infor- follow. mation are given in the appendix.

It will be noted that certain of these documents provide extensive additional bibliographies while others do not. The Focus Focus Group on Karst Hydrology

Focus Group on Karst Hydrology – Conceptual Models, Aquifer Characterization, and Numerical Modeling

Group Participants: Martin Sauter, Leader. Matt Covington, Lee Florea, Franci Gabrovsek, Yongli Gao, Ronald Green, Jason Gulley, Russell Harmon, Ellen Herman, Pierre-Yves Jeannin, William K. Jones, Todd Kincaid, P.J. Moore, John Mylroie, Ira D. Sasowsky, Elizabeth Screaton, and Carol M. Wicks

INTRODUCTION CONCEPTUAL MODELS OF WATER FLOW IN KARST SYSTEMS low of water in a karst catchment is mainly determined by the hydraulic gradient to a point of discharge (spring, In recent decades, it has been realized that karst processes must Friver or coastline), the geometry of the karst features be considered in a broader context than the traditional dissolu- (fissures, fractures, conduits, and other zones of dissolution- tion in circulating meteoric water. Karst, both surface land- enhanced high hydraulic conductivity), the sources (sinking forms and caves, formed by water circulating downward (and streams, sinkholes, and the epikarst), and temporal variation also laterally) from a meteoric source on the land surface is of the recharge input. While the bulk hydraulic gradient can referred to as epigenetic karst. Karst, mostly caves, formed be determined in the field, the geometry of the karst features, by water migrating upward from depth is referred to as hypo- their hydraulic parameters and their spatial distribution require genetic karst. Hypogenetic karst is often the product of dis- considerable effort to quantify. Although some caves appear solutional processes beyond the carbonic acid chemistry that very spectacular and voluminous (if accessible), the fraction is the primary driver for epigenetic karst, especially sulfur and of rock occupied by caves in a karst system frequently is less sulfuric acid chemistry. Water often flows in epigenic karst than a few percent. This implies that these features are difficult aquifers in a turbulent regime. Flow in hypogenic systems is to detect by drilling and even more difficult to parameterise mostly laminar with caves embedded in a diffuse flow sys- hydraulically. Due to their important role in conducting water tem that is completely decoupled from the surface hydrology. flow, both vertically in the vadose zone and horizontally in the Models for karst aquifers developed in diagenetically mature, phreatic zone, they cannot be neglected either (Klimchouk et well-compacted carbonate rocks usually need to take account al., 2000). only of the conduit permeability and the fracture permeability. Matrix permeability is often very low although it is sometimes In order to be able to predict flow through karst aquifers, it is substantial. Such aquifers are referred to as telogenetic karst. critical to first design plausible conceptual models. Concep- Other aquifers develop is young carbonates where diagenetic tual models provide a framework to support more quantitative processes may be incomplete. In eogenetic karst aquifers, ma- mathematical models. Conceptual models are followed by a trix permeability is usually a dominant part of the flow system. first principles understanding of ground water flow including Fractures may be a relatively minor feature. quantized inputs and outputs. The final step is the quantitative mathematical or computer model. Such models allow, for Conceptual Models for Telogenetic Karst example, the assessment of water resources, vulnerability of Aquifers karst ground water to contaminants, potential flooding risks, and infiltration rates into caves. Most conceptual models of water flow in telogenetic karst distinguish three main zones (compartments) in the vertical direc- In this report, different conceptualizations of karst systems are tion. These are: the soil zone and epikarst, the unsaturated or provided in order to convey an understanding of the different vadose zone, and the phreatic zone. Although most conceptual recommendations formulated for future research initiatives. models include similar structural features, the flow and storage The recommendation categories are assembled into differ- processes assigned to them display large variations. Figure 1 ent groups: processes, quantification of recharge and infiltra- presents a conceptual model of such a karst system. tion, characterization, models, and cross-disciplinary research needs. The epikarst, a zone of increased weathering near the land surface, determines the distribution of recharge to a karst aquifer

Frontiers of Karst Research 77

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Focus Group on Karst Hydrology

Figure 1. Conceptual model of a telogenic karst system in dense well-lithified carbonate rock (After Liedl and Sauter, 2003).

in both space and time. It can be visualised as a perched aquifer medium the resulting hydraulic parameters are difficult to inter- system channelling diffuse input towards shafts and sinkholes. pret from standard investigation techniques such as hydraulic The presence of the epikarst is believed to explain the highly tests, and cannot be easily regionalized at the catchment scale. heterogeneous water input to the system, both in space and time. Furthermore, it links climatic and near-surface geologi- Conceptual Models for Eogenetic Karst cal conditions with the karstification of a limestone aquifer, de- Aquifers fining both the hydraulic and the chemical boundary conditions for the development of the karst system. An understanding of Eogenetic aquifers, almost by definition, have high matrix the functioning of the epikarst is therefore a prerequisite for permeabilities. The high matrix permeability creates a large the quantification of infiltration (Jones et al., 2003; Geyer et accessible storage that is a significant contribution to the flow al., 2008). system. Flow in the matrix can be modelled as classical Dar- cian flow. However, it is the difficult to account for extensive Due to the complex and heterogeneous nature of karst systems, conduit development in such aquifers, but the conduits are quantification of recharge input is a major challenge. In or- definitely present. Quantitative models must then address the der to be able to predict short term responses of karst systems interchange of conduit flow and matrix flow. Aquifers in the to recharge events with numerical models, knowledge of the Paleozoic rocks of eastern United States are telogenetic karst quantity of recharge, it’s temporal and spatial variability and of as are many of the aquifers of Europe and the Mediterranean. the infiltration mechanism is a prerequisite. Examples of Eogenetic karst are the Floridan Aquifer and the carbonate islands of the Caribbean. Karst aquifers are characterized by highly varied hydraulic properties which are a result of the complex interactions be- In coastal regions, where carbonate rocks extend below sea tween karst conduits, discrete fractures and the rock matrix. level a circulating pattern of sea water and fresh water develops Conduits are characterized by low storage and high flow ve- which enhances karst development in both dense and diage- locities, while the discrete fissured system and the rock matrix netically immature limestones. The resulting cavities, known display much higher storage and low flow velocities. Due to as halocline or flank margin caves are thought to contribute to this dual-porosity, dual-permeability structure of the carbonate reservoir porosity later in their geologic history (Sasowsky et

78 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Focus Group on Karst Hydrology

al., 2008). Flank margin caves have an entirely different role is relevant. For example: The increase of discharge at springs in aquifer hydrology than do the integrated conduit systems of can be attributed to: a) the specifics of storm intensity, spatial telogenic karst. distribution and temporal variability, b) the characteristics of evapotranspiration at the soil – vegetation – atmosphere bound- Modeling efforts on flank margin caves has been undertaken ary, c) the heterogeneous infiltration process with rapid infiltra- but much remains to be done (Fig. 2) tion via fractures and sinkholes and slow infiltration through the vadose zone rock matrix, and d) the dualistic flow and storage processes in the phreatic zone (see also Smart and Hobbs, 1986) (Fig. 3). The analysis of spring discharge data therefore requires an appropriate conceptual model, the identification of the individual flow processes in the different compartments, as well as characterization of the geometry and hydraulic properties of the flow paths. Continuous measurement of the input functions such as precipitation and evapotranspiration is important.

Figure 2. Computer generated model of a flank margin cave (box pattern) compared with a cave survey (smooth pattern). From Labourdette et al. (2007).

Conceptual Models for Hypogenetic Karst Systems

Karstification by rising deep-seated solutions has been recognized so recently that the development of conceptual models remains an open problem (Klimchouk, 2007). Unlike both telogenetic and eogenetic varieties of epigenetic karst where Figure 3. Schematic diagram to demonstrate the superposi- tion of different processes and the influence of the different the chemical processes are moderately well understood, even compartments on spring discharge (after Smart and Hobbs, the chemistry of hypogenetic karst poses many open questions. 1986) Some of the deep-seated solutions are derived from petroleum reservoirs, some are associated with volcanic activity, and some are deep-seated brines. For the most part, hypogenic karst flow Further process understanding is required for the quantification systems are recognized only by caves that are later exposed. of: Active systems can be intercepted by drill holes, but both chemistry and flow hydraulics of the processes taking place at a) Unsaturated flow: Are film flow processes relevant? depth are poorly understood. b) Open channel flow: Is this process relevant at various flow PROCESSES conditions? Can certain observed thresholds, observed in the discharge pattern be explained by open channel flow condi- A number of relevant flow, transport, to some extent reaction tions? processes, have been implemented into the various types of models (discrete, hybrid, continuum models). c) Turbulent flow: Although difficult to parameterize, turbulent flow processes can explain many observations in karst flow In a karst system, a number of different processes are superim- and transport (e.g. dispersion, tailings in tracer breakthrough posed so that the influence of individual processes is difficult curves). A quantitative and efficient approach in the identifica- to quantify, especially if their spatial and temporal variability tion and model implementation of turbulent flow processes is

Frontiers of Karst Research 79

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Focus Group on Karst Hydrology

required. (Computational fluid dynamics models can simulate a) The quantification of total recharge at the soil base level: theses processes at laboratory scale, but the calculations are Soils in karst regions are often very thin and frequently very much more difficult at the catchment scale). patchy (often alternating between bare rock and soil pockets). Therefore, traditional classic soil moisture balance techniques d) Sediment transport: What determines sediment transport; do not necessary apply. Methods need to be developed, includ- are there thresholds? ing hydrological, remote sensing and geophysical techniques, which take into account the specifics of karst soils. e) Variable density flow: Flow in island and coastal karst settings is determined, apart from the general characteristics of b) Characterization of flow and transport in the vadose the carbonate limestone materials, by the variability in salt con- zone: The vadose zone, extending between the base of the centrations. Although in many circumstances, advective flow epikarst and the ground water table frequently measures more dominates flow behaviour there might be still-water environ- than 50 metres, sometimes several hundreds of metres. The ments, where convective processes might be important (e.g. for presence of this zone transforms the already complex recharge carbonate dissolution processes etc.). input signal into an even more complex one, composed of a rapid and a slow input with variable fractions and a variable f) Multiphase flow: Although important in the context of e.g. temporal response. An integrated approach, using hydrogeo- contaminant transport, multiphase flow processes will be dif- logical and geophysical characterization techniques is expected ficult to identify and parameterise in this complex and highly to provide at least some indications of the respective input heterogeneous environment. However, the development of functions. models that incorporate multiphase flow processes (also air / water) might assist in understanding certain observations. These types of investigations are expected to be most successful if performed on a catchment scale. Quantitative measure- g) Conduit-matrix exchange, and the geochemical reactions ment of input (precipitation) and output (spring discharge) can that occur across this boundary, particularly in eogenetic karst. then be measured and an overall water balance established.

Apart from the identification of processes and the characterisa- CHARACTERIZATION tion of process parameters, it is also important to initiate investigations into an efficient implementation of these processes The prime characteristic of telogenetic karst aquifers is, apart into the respective (discrete, hybrid, continuum) numerical from their genetic history, the heterogeneity and the extreme models. To improve applicability of models to a larger segment contrast in the hydraulic parameters of highly conductive, but of problems and the overall quality of model input, we need to low storage, conduits and the low conductive but high storage improve our understanding of measurable parameters in karst fractured matrix. Although the conduit network dominates the systems and develop basic empirical relationships based on flow pattern, it is extremely difficult to detect with drilling or those parameters. traditional geophysical techniques because of its low volume fraction in the bulk aquifer. Therefore, active conduits cannot Badly needed are investigations into the system responses and be properly characterized hydraulically or structurally, unless thresholds. Extreme storm events are useful probes if aquifers they are accessible for divers or located within the vadose por- are adequately instrumented so that complete response records tion of the cave system. Indeed, a fundamental limitation of all can be obtained. modelling efforts to date is that the conduit system must be put into the model “by hand”. If the location and characteristics of QUANTIFICATION OF RECHARGE INPUT the conduit system are completely unknown, any models and AND INFILTRATION the corresponding model output will be of limited value.

Independent assessment of recharge is of prime importance Techniques need to be developed that provide the basis to lo- for any model construction. Since recharge in karst aquifers cate individual conduits via novel geophysical or telemetric occurs both spatially distributed (diffuse) as well as locally techniques and detailed mapping of hydraulic potential. Fur- concentrated via sinking streams, shafts and dolines (discrete), thermore, cave mapping can be expected to provide analogues substantial effort needs to be invested in its characterization. for submerged cave networks and the understanding of the ge- This includes: netic history of karst networks (geomorphological, paleokarst investigations) as well as numerical karst genesis modelling

80 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Focus Group on Karst Hydrology

can supply additional information for the outline of the geom- d) Identify suitable benchmark sites (karst catchments) for etry of conduit networks. common interdisciplinary experiments. These catchments will: In the case of hypogenic caves, the mere existence of a cave is difficult to determine as discrete recharge and discharge points i. Have long-term data records, are not available for sampling and measurement. Such caves ii. Be relatively “simple” geologically and hydraulically can be quite large, for example, Carlsbad Caverns, yet have no iii. Provide the basis for multidisciplinary research topics, e.g. obvious indication on the surface. Geophysical methods would the influence of microbial and biogeochemical processes on also be useful, particularly if they could detect cavities at great carbonate dissolution depths. iv. Be suitable, so that jointly (hydro-bio-chem) organised experiments have a chance to alter system behaviour. MODELS AND MODEL DEVELOPMENT CONCLUDING REMARKS A variety of modelling strategies have been developed. All show some promise but none are really satisfactory. The re- Although an obvious statement, it is important to reiterate that cent employment of lattice Boltzmann techniques for process model selection needs to be based on the formulation of the modelling in karst appears promising. The overall goal should problem to be solved. This requires a concise statement of the be process-oriented models that emphasize efficient numerical problem, an initial plausible conceptual model, the identifica- computation of large, catchment scale models, while honouring tion of relevant processes and parameterisation strategies. small scale heterogeneities. Since wrong conclusions are frequently being drawn from erro- Furthermore, since the different models require different model neous model selection and interpretation of model results, it is parameters (e.g. spatially averaged, discrete geometric, double suggested that a tool box/guidelines be developed for of sound continuum etc.) model adapted characterisation strategies need karst modelling protocols. This means that guidelines should to be developed. The differences in parameterization strategies be formulated that suggest investigations and modelling strate- need to be highlighted and the relationship between actually gies, based on the karst conceptual models and the problem in measurable parameter (via hydraulic tests, etc.) and model cali- question. Benchmark catchments and sample calculations can brated parameters needs to be clearly stated. demonstrate their application to real case studies.

COMMUNICATION AND Quantitative assessment of flow and transport in karst, the iden- CROSS-DISCIPLINARY RESEARCH tification of important processes as well as the determination of process parameters among ground water professionals is gener- In order to further research into the hydrology, biology, ecol- ally hampered by the lack of understanding of the particulars ogy and geochemistry of karst, appropriate platforms need to of karst systems. It is important to convey to the professional be set up: public via appropriate media techniques (karst portal, etc.) that “karst is different.” It is also important to convey the expanded a) Provide a common language and understanding. concepts with the distinction between epigenetic and hypoge- b) Organize joint experiments. netic karst and between telogenetic and eogenetic karst. c) Provide common data bases (cave surveys, tracer experiments, hydraulic tests).

Frontiers of Karst Research 81

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Focus Group on Geochemistry and Climate

Focus Group on Geochemistry and Climate

Group Participants: Jay Banner, Leader. Penelope Boston, Liza Colucci, Brian Cowan, Amy Frappier, Cara Gentry, Russell S. Harmon, Brian Katz, Andrew Long, Jonathan B. Martin, MaryLynn Musgrove, Jud Partin, Jessica Rasmussen, Corinne Wong and William B. White

GEOCHEMISTRY OF MODERN KARST Challenges to Water Resource Sustainability SYSTEMS Understanding of flow through karst aquifers, and the chemical Importance to Science and Society and isotopic composition of the water clearly have high social relevance because of the susceptibility of karst to contamina- ater use management and water resource conservation through rapid infiltration (e.g., Field, 1988; Zuber and tion practices require knowledge of aquifer func- Motyka, 1994; Boyer and Pasqarell, 1995; Vaute et al., 1997), Wtioning and processes. This is particularly true for and the impacts of karst processes on water resources. Fun- karst groundwater resources, which are of enormous global damental karst processes such as karstification, speleogenesis, economic value. Karst groundwater resources supply drinking aquifer evolution, and sinkhole formation are all controlled by water to an estimated 25% of the world’s population (Ford and geochemical reactions (Dreybrodt, 1981; 1988; Beck, 1986; Williams, 1989). In the United States, karst covers approxi- Palmer, 1991; Romanov et al., 2003; Gabrovsek et al., 2004). mately 20% of the land surface (Fig. 4) and provides substan- Detailed chemical and isotopic compositions of karst waters tial amounts of the groundwater used for drinking water. Karst and their variability in time and space could provide additional aquifers and associated springs provide water for agriculture insights into flow and reaction in karst aquifers. Challenges to and aquiculture, drinking water, recreational and tourist oppor- water resource sustainability with respect to karst are numer- tunities, and habitat for myriad aquatic and terrestrial organ- ous and are complicated by the inherent temporal variability isms. of karst aquifers. Challenges include drought and flood vari-

Figure 4. Karst areas of the U.S. Map provided by G. Veni and reproduced with permission of the American Geological Institute.

82 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Focus Group on Geochemistry and Climate

Figure 5. Schematic diagrammatic representation of the coupling of flow in conduits and through matrix porosity at A. Base flow conditions, B. Flood conditions. (From Martin and Screaton, 2001). ability, residence time variability, aquifer contamination, and zoa, bacteria), suspended sediment and turbidity, and pharma- increasing demand for water resources. ceuticals (Carey and Lloyd, 1985; Boyer and Pasquarell, 1996; Katz et al., 2001; Barrett and Williams, 1989; Pasquarell and Unique Vulnerability of Karst Systems to Boyer, 1996; Mahler and Massei, 2007; Wolfe and Williams, Contamination 1999; Mahler et al., 2000; Dussart et al., 2003; Kuczynskaa et al., 2003; O'Reilly et al., 2007; Ryan and Meiman, 1996; Karst aquifers are highly productive sources of water, but many Massei et al., 2003; Drysdale et al., 2004; Wicks et al., 2004; of the unique characteristics that render them productive also Mahler, et al., 2006). make them extremely vulnerable to contamination (Vesper et al., 2001). Overlying soils are frequently thin to non-existent, KEY SCIENCE QUESTIONS providing little opportunity for filtration, sorption, or bacterial remediation of contaminants. Much of the infiltration is fo- What is the extent, interconnectedness and cused through discrete recharge features such as swallow holes structure of conduit vs. diffuse flow through and open fractures, which transmit water rapidly and provide karst aquifers? minimal filtration. Within the aquifer, most of the transport occurs within a network of fractures and conduits. As a result, One of the key challenges facing karst science is how to better flow velocities are typically orders of magnitude higher than in understand and describe the extent and interconnectedness of porous media aquifers, sometimes as high as several kilome- conduit and diffuse flow. Karst aquifers can be considered to ters per day. As a result of rapid flow velocities, ground-water have dual or even triple porosity - a low permeability matrix, flow can become turbulent, entraining particles and associated in which most of the storage occurs, and a conduit system, in contaminants within the aquifer, and increasing turbidity and which most of the transport occurs. A third type of porosity, transport of hydrophobic contaminants, including pathogens. micro-fractures with permeability intermediate between the The rapid transport of water and contaminants from the surface matrix and the conduits, is often assumed as well. through the aquifer to springs and wells may leave little time for response and/or remediation. Additionally, the extreme isot- Aquifer recharge can occur as focused recharge through dis- ropy and heterogeneity of karst systems complicates prediction crete features such as swallow holes or open fractures, or as of transport velocities and even direction of flow, and leaves diffuse recharge through soils into the matrix. When no re- traditional contaminant transport models inadequate to address charge is occurring, water stored in the matrix diffuses into the these unique vulnerabilities and complexities. Numerous hu- conduit system; when focused recharge occurs, higher head in man activities, including urbanization, agriculture, waste-water the conduit system may result in diffusion from the conduits discharge, and deforestation have resulted in a variety of an- into the matrix. Thus most of the transport occurs through the thropogenic changes and contamination to karst aquifers, in- conduit network, but under some conditions it is matrix water cluding nitrates and other nutrients, pesticides, solvents and that is being transported, and under others it is surface water. other volatile organic compounds, pathogens (viruses, proto- Understanding of the different compartments of karst aqui-

Frontiers of Karst Research 83

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Focus Group on Geochemistry and Climate

fers (e.g. conduit, fracture and intergranular/matrix porosity) tions of microbial communities in caves. These microbes may has been improved through developments of deterministic and also influence the chemical and isotopic compositions of karst stochastic models of the aquifers. Deterministic models com- waters. Fractionation of carbon isotopes has been used as evi- monly use physical measurements (head, flow, temperature); dence of remineralization of organic carbon and the geochemi- some geochemical characteristics (e.g. specific conductance) cal pathways of the microbial communities (McMahon and can also provide constraints on models Chapelle, 1991). These pathways require terminal electron acceptors, which are commonly oxygen in karst systems because The exchange and movement of water via conduit and diffuse of the oxygen-rich atmospheres of many cave environments. flow plays a crucial role in determining aquifer recharge and Many other terminal electrons can be important, however, and storage, aquifer yield, and contaminant transport and storage there is recent evidence of microbially-mediated sulfur reac- processes in karst, and has significant bearing on understanding tions in caves (Engel et al., 2007). Additional terminal electron and modeling the hydrology of karst systems. Our understand- acceptors include metals such as iron and manganese. Oxida- ing of the specifics and variability of conduit vs. diffuse flow tion-reduction (redox) reactions provide intermediate amounts in karst aquifers remains a significant topic of study, which is of energy to carbon-oxidizing microbes and are likely to take crucial for addressing quantification of the extent of conduit place on a continuum of reactions between oxygen and sulfate vs. diffuse flow, the controls of flow on the geochemical evolu- and carbonate reduction (Froehlich, 1979). Cycling of metals tion of karst waters, differences in flow between the vadose and is readily observed as coatings of iron and manganese on cave phreatic zones, temporal variability in vadose vs. diffuse flow surfaces (e.g. goethite, limonite, and birnessite) and their pre- under conditions such as storms, and aquifer response to stress cipitation can incorporate a range of other elements (Ba, Sr, P), (such as pumping). thereby impacting the overall chemical composition of karst waters. The concentration of oxygen and other terminal elec- What are the Mechanisms That Control the tron acceptors may range widely throughout an aquifer, par- Chemical Evolution of Water Along a Flow ticularly outside of cave environments. Redox reactions may Path From Surface Water to Ground Water also be important in microenvironments of the epikarst and in and Return to the Surface? matrix fracture and primary microporosity, where water can be separated from the atmosphere, allowing oxygen to be quickly The varieties of flow paths that characterize karst aquifers cre- consumed. Focused geochemical studies of these environ- ate a range of residence times for water from hours (Martin and ments and water associated with different karst conditions may Dean, 2001) to decades (Long and Putnam, 2004; 2006). These provide for improved understanding of the role of microbial differences in residence times influence the chemical evolution processes. of water, both in the water’s major element chemistry through carbonate mineral reactions (e.g. Plummer, 1977), and also in What are the Impacts of Land Use Change its minor and trace element concentrations and isotopic compo- and Climate Change on Water Resources sitions. In addition, contaminants introduced from recharge of and Can the Impacts Be Distinguished? surface water will react at varying rates and magnitudes with the rocks and surfaces of the aquifer materials, depending on Land use changes have significant impacts on karst areas. An- their physical characteristics and mineralogy. For example, thropogenic activities such as urbanization and development, phosphorous can be a limiting nutrient for many karst spring deforestation, agriculture, ranching, livestock grazing, fire sup- systems, but also can take part in surface reactions with car- pression practices, urban landscaping, landfills, waste-water bonate minerals and metal oxides. In contrast, nitrate, a widely discharge, sewage disposal and sewage and municipal water recognized contaminant in karst systems because of the wide- infrastructure can all change the nature of land surface and spread application of fertilizers, undergoes very few reactions water resource interaction (e.g., Boyer and Pasquarell, 1999; involving aquifer rocks. Drew, 1996; Garcia-Fresca, et al., 2004; Harding and Ford, 1993; Parise and Pascali, 2003; Sauro, 1993; Wang et al., 2004; What is the Fate and Transport of Redox- Williams, 1993). Negative impacts such as soil degradation, Sensitive Elements and Their Microbial groundwater salinization, increases in contaminant concentra- Consequences and Feedbacks? tions, degradation of water quality, ecosystem and biodiversity loss, and changes in aquifer storage, recharge, and water The role of microbes in geochemical functioning of karst aqui- availability can result in karst regions as a result of land use fers is becoming increasingly clear through DNA characteriza- changes.

84 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Focus Group on Geochemistry and Climate

The impact of climate change on karst and karst resources represents an emerging multifaceted question. Climate change may significantly affect precipitation amounts and seasonality, water availability, river runoff and streamflow, and frequency and intensity of droughts and floods (IPCC, 1997, 2001, 2007). The intersection of anticipated climate change impacts and the inherent temporal variability of karst water resources necessi- tate a better understanding of climate change impacts in karst regions. There is, however, a lack of information on 1) the mechanisms and timescales at which land use changes and climate change impact karst aquifers, 2) if these changes are linear or threshold induced, and 3) possibilities of reversibility and/or Figure 6. Rapid changes in specific conductance (solid line) remediation of such changes. This information is essential for and atrazine concentration (diamonds) reflect the influx of water resource management and planning, and for establishing surface water to Upper Barton Spring, Austin, TX, following a storm. Adapted from Mahler and Massei (2007). sustainable management practices. Effects of land use change and climate change on water resources may be synergistic, but Because of the extreme heterogeneity of karst systems, data measuring and distinguishing them presents a significant chal- collected at high spatial resolution is valuable. In general, lenge. Paired or comparative studies of watersheds, regions, monitoring of springs should be considered a first priority, as or aquifers may provide insight, as may predictive forecasting. springs act as integrators of the entire karst system (Quinlan, The integration of hydrologic and regional scale climate mod- 1989). Even wells within a few meters of one another can show els is a developing research direction (Gulden et al., 2007; Niu significantly different responses in piezometric level and geo- et al., 2007) that will have direct application to karst systems. chemistry (e.g., Mahler et al., 2000). This piezometric instability, in turn, can be extremely useful in investigating the spatial HOW TO ADDRESS THESE KEY SCIENCE heterogeneity inherent in karst (Palmer, 2002). QUESTIONS Integrate Geochemical and Physical High-Resolution Temporal and Spatial Information to Enhance Current Modeling Analysis of Geochemical and Physical Capabilities Variability, Particularly for Extreme Events (Floods, Droughts, etc.). Understanding the complexity of karst geochemical and physical dynamics requires the use of multiple integrated tech- Information gained from high resolution temporal and spatial niques. Modeling geochemical and hydrologic signals and analysis of geochemical and physical variability, particularly responses is particularly effective in karst aquifers through ap- for transient events such as floods, will be extremely valuable plications such as lumped-parameter modeling (deconvolution) for improving our understanding of karst systems. Numerous and frequency or wavelet analyses (Long and Putnam, 2004; studies have demonstrated that following rainfall, large changes Massei et al., 2006). These methods are useful for understand- in water level, spring discharge, turbidity, and concentrations of ing and quantifying geochemical and physical processes of karst natural and anthropogenic chemical constituents may occur at a systems, particularly those related to conduit and diffuse flow. time scale of hours (Fig. 6) (Ryan and Meiman, 1996; Mahler, Recently developed models for interpreting age-dating tracers 1997; Mahler et al., 2000; Massei et al., 2002). For example, (CFCs, tritium, SF6, 85Kr, 36Cl) for karst aquifers are useful concentrations of atrazine in a karst spring in Austin, Texas, for estimating the residence times of conduit and diffuse flow increased by a factor of more than 50 in the 20 hours following and the characteristic responses of these distinct components rainfall (Mahler and Van Metre, 2000). Thus, monitoring and (Long and Putnam, 2006). These methods could be enhanced models designed for time steps of months or even weeks are further by incorporating a 14C data component. Application of likely to fail to detect the dynamic responses of karst systems. specialized tools such as these with common geochemical mod- Analysis of breakthrough curves can provide information such els such as NETPATH and PHREEQC (Plummer et al., 1992; as transport times, dilution factors, location of contaminant Parkhurst and Appelo, 1999) could result in robust approaches sources, and diffusion factors, and are critical for calibrating to understanding karst processes. Modeling geochemical pro- transport models. cesses in conjunction with fluid flow is a frontier in hydrologic modeling and would be particularly useful when considering

Frontiers of Karst Research 85

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Focus Group on Geochemistry and Climate

karst dynamics. Application of advanced parameter estimation infection with chlorine). Understanding the geochemical fac- methods (Doherty, 2002) can be used to further the science of tors involved in the degradation or alteration of NOM in the modeling karst aquifers through more efficient data processing environment is critical for accurate prediction of the formation and objective quantification of aquifer properties. of DBPs, such as trihalomethanes and other contaminants (e.g., Rostad et al., 2000). Apply Newly-Developed Techniques and Explore Additional Applications of Existing GAPS IN THE PRESENT RESEARCH Techniques. INFRASTRUCTURE

The integration of multiple tracers in an aquifer system may Responses in a karst system can be long term and short term. provide more robust and/or complementary information regard- Long-term (years to centuries) and short-term (hours to weeks) ing tracer behavior and geochemical and aquifer processes. A time-series data are useful for understanding these varied re- large number of potential anthropogenic and natural tracers are sponses. Such data can be acquired if funding priority is given available to assess mechanisms of groundwater geochemical to establishing intensive monitoring networks in karst sys- evolution; many are underdeveloped or poorly understood due tems. Long-term data collection efforts for geochemical trac- to complexities in tracer behavior. Natural geochemical tracers should be conducted similar to the way hydraulic data is ers such as trace element ratios and stable isotopes historically typically collected (e.g., streamflow, spring flow, and hydraulic have been applied to address karst groundwater evolution pro- head). NSF’s proposed WATERS Network would help achieve cesses, mineral solution equilibria, ground water flowpaths, sur- this goal. Furthermore, analysis of the geochemical response of face water – ground water interactions, hydrograph separation, karst systems to changes in the landscape and ecosystems that and sediment transport. Anthropogenic contaminants in karst are developed in karst terrains would permit a deeper process- systems provide opportunities to develop new tracers that may level understanding of the functioning of karst hydrology. To provide specific details of contaminant transport (e.g., Mahler acquire these data would require an intensive monitoring sys- and Massei, 2007). Novel applications of radiogenic isotopes, tem such as that proposed for the National Ecological Observa- such as Sr isotopes (Musgrove and Banner, 2004; DeMott et tory Network. Analytical advances in terms of isotopic analy- al., 2006) may provide insight into processes of soil contribu- sis of specific organic compounds in water, portable chemical tions to groundwater geochemistry, land use and urbanization and isotopic analysis via methods such as Raman spectroscopy impacts in water resources, groundwater mixing processes, and and mini-differential optical absorption spectrometry will be sources of dissolved constituents to groundwaters. Soil profiles required for the next step in in situ monitoring of karst flow of chloride and tritium can be used to estimate past recharge systems. Small, secure, mobile, inexpensive, and durable in- rates over tens and, in some cases, hundreds of years (Herczeg strumentation to automatically collect a range of geochemical and Edmunds, 1999) with implications for water balance ques- and physical data in multiple aquifer components (e.g., soil, tions and resource sustainability. Applications of tracers such vadose, phreatic, epikarst, conduit porosity, diffuse porosity, as radiogenic isotopes, rare-earth elements, Fe isotopes, and along flowpaths) at high resolution will be a critical step in ad- biomarkers are frontiers in karst geochemistry. dressing research needs.

The vulnerability of karst aquifers to contamination is exac- As karst science grows into an integrated field of hydrologic, erbated by recharge of surface water into underlying aquifers modeling, geochemical, biological, and ecosystem disciplines, (Katz et al., 1995). These processes can result in groundwater- a need exists for storing and sharing of geochemical and physi- quality problems, such as high concentrations of iron and hy- cal data from karst systems. A data portal, or central system for drogen sulfide, and undesirable bacteria, protozoa, and fungi data information sharing and discussion, beyond the published (Krause, 1979; McConnell and Hacke, 1993; Katz et al., 1995; scientific literature was recognized as a necessity for future Plummer et al., 1998). The combined use of various geochemi- integrative efforts in karst science in order to address larger cal tracers, including radiogenic and stable isotopes, tannic collective goals. Additionally, beyond the karst community, a acid, chloride, and silica have been effective in calculating the need for improved cross-disciplinary communication and col- proportion of river water that mixed with groundwater (Katz laboration with the broader geochemistry and climate com- et al., 1997, 1998). Blackwater streams are common in the munity exists, to include scientists with expertise in ecosystem southeastern United States and can introduce large amounts of function, global atmospheric and marine and climate systems. natural organic matter (NOM) that can produce disinfection Improved interaction and communication with the public is a byproducts (DBPs) during drinking water treatment (e.g. dis- continuing need.

86 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Focus Group on Geochemistry and Climate

CLIMATE CHANGE AND KARST ocean circulation, to monsoonal and El Nino-Southern Oscilla- SYSTEMS tion phenomena. One example of this potential is a speleothem record from Hulu Cave, China, which shows clear connec- The observed warming trend over the last century has moti- tions between atmospheric greenhouse gas concentrations and vated extensive research of the interactions between temper- temperatures, as determined from ice core records, and Asian ature and precipitation variability in past, present and future monsoon intensity as reflected in speleothem oxygen isotope climates. As detailed in the 2007 IPCC report, understanding variations (Wang et al., 2001). where, when and why dramatic variability occurs is critical to societies managing food and water resources in climatically What are Feedbacks Between the Global marginal regions, such as northern China and the African Sahel Carbon Cycle and Karst Systems? (Menon et al., 2002; Nicholson et al., 1998). Karst system science can offer a unique perspective on the temporal and spatial Carbonate rocks are a relatively soluble and large source of car- links between climate and hydrology. Given that karst ground bon that constitutes a significant component of the global bio- water is an important component of the world’s drinking water, geochemical cycle of carbon. Understanding the role of karst karst systems are of central importance to water quality and in the global carbon cycle will require studies of the feedbacks water availability issues. between changes in sea level with changes in exposure, weathering, and production of calcium carbonate in coastal regions. CLIMATE CHANGE – KEY QUESTIONS Studies will be needed that can address the timescales on which carbon is sequestered and released in terrestrial systems. In What is the Relationship Between Climate Change and Water particular, these studies will need to address the question “To

Availability? what extent are karst systems a source or sink for CO2?” Key constraints on karst-carbon cycle feedbacks will come from an The mechanisms driving high-frequency paleoclimate vari- improved understanding of the rates of karst geomorphic devel- ability in climatically-marginal regions are not well known due opment relative to temporal changes in sea level, atmospheric to the general scarcity of long-term, high-resolution, precisely composition and rainfall amounts. dateable proxies for precipitation in continental interiors. Spe- leothems have the potential to provide this information through What are Feedbacks Between the Local multiple proxies including geochemical data, physical proxy Carbon Cycle and Karst Systems? data such as annual banding, and growth rate (Musgrove et al., 2001; Asmerom and Polyak, 2004). Given the large spatial and Carbon cycling at the landscape level may be significantly af- temporal distribution of speleothems, these studies can provide fected by interactions between the surface and underlying karst. unprecedented insight to: The transport of carbon out of soils and into the karst environment is a flux that is rarely treated in carbon-cycle modeling.

• The underlying mechanisms that have modulated past cli- In the opposite direction, the movement of CO2 from caves, mate variability across multiple timescales, from seasonal springs and wells has not been documented in a comprehensive to millennial (Asmerom et al., 2007). manner at the landscape scale. This flux is also important in • The frequency and magnitude of extreme events such as modeling speleogenesis and subterranean calcite deposition. droughts, floods, and cyclones (Frappieret al., 2007). • The implications of past variability on future climate HOW TO ADDRESS THESE KEY SCIENCE change. QUESTIONS

How is Climate Variability Expressed in Ter- Many of the questions outlined above can be addressed through restrial Environments? the development and integration of speleothem records that are well-dated, employ multiple proxies, and come from a wide The pursuit of paleoclimate reconstruction has centered on ex- range of geographic locations that span multiple scales. This cellent marine and ice core records, there are significant gaps in may be similar to the approach of constructing and compil- our understanding of climate change in terrestrial environments. ing marine and lake records to produce the CLIMAP data set. Speleothems can provide significant advances in this area, both Significant resources would be required to build a “SPELEO- in terms of relatively long-term, astronomically forced chang- MAP” data set. es, to abrupt climate change, such as those driven by changes in

Frontiers of Karst Research 87

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Focus Group on Geochemistry and Climate

Essential to reconstruction of past climate change using spele- ter target specific samples while preserving cave formations. othems is a rigorous evaluation of and calibration of what the Minimally-invasive sampling can be done via small cores to elemental, isotopic, and growth rate proxies represent. Three determine age ranges and suitability for geochronology of main approaches may be taken toward this end: 1) comparison multiple speleothems without removing them from the cave of speleothem proxies to multiple independent records of cli- environment. Samples may also be imaged by computed X- mate change over the same time interval in the same region; 2) ray tomography to determine if they are petrologically suitable, growth of speleothems under controlled laboratory conditions and then replaced in the cave environment if deemed unsuitable in which a range of parameters can be varied; and 3) growth of for further study (Mickler et al., 2004b). At present, funding speleothems in modern karst environments in which environ- agencies will typically not fund projects of this design because mental parameters can be monitored. of the length of time they would cover. A renewed emphasis on small grants reserved for exploratory, or reconnaissance, re- GAPS IN THE PRESENT RESEARCH search would help promote minimally invasive sampling. Also, INFRASTRUCTURE a central storage facility of samples, either virtual or real, such as those already employed by the ice core and sediment core Network of Intensive Monitoring of Modern communities, can help to minimize sample collection where System for Assessment and Calibration of research group interests overlap. To obtain a sample from this Speleothem Proxies facility, a research group would need to write a short proposal about the project and how the samples can address those prob- Following on the discussion in the section above, the evalu- lems raised. Lastly, a central archive of not only proxy data ation and calibration of speleothem proxies must include the collected, but metadata about sample selection processes as assessment of equilibrium precipitation of speleothem calcite well, will help future studies produce better quality datasets. (Mickler et al., 2004, Banner et al., 2007). The most unequivo- Replicate records in multiple samples from the same cave as cal means to achieve this assessment is through the analysis of well as in caves across a region would help validate and fa- the modern system, which can be accomplished through estab- cilitate interpretations of karst paleoclimate reconstructions lishing monitoring networks in karst aquifers. Understanding (Musgrove et al., 2001; Wang et al., 2001; Treble et al., 2005; of temporal changes of the modern landscape above karst aqui- Williams et al., 2005; Johnson et al., 2006; Wang et al., 2006). fers and the modern aquifer system are going to be essential This helps to separate hydrologic processes from climatic pro- for interpreting seasonal to millennial scale records in the past. cesses and strengthens climatic interpretations drawn from the This will require studies of changes 1) above caves (thickness, records. While this may introduce an initial higher cost, it will moisture, water chemistry, productivity of soils, vegetation, and reduce final costs by producing reliable records which do not weather, 2) at cave drip sites (cave-air meteorology, drip rate, need replication. A better spatial coverage of speleothem re- drip composition, and 3) event sampling. Studies of human- constructions can also addresses local versus regional-scale cli- driven land-use change in karst systems will make speleothem mate changes, much like modern satellite datasets. studies more relevant to human time scales. This will require an even more extensive monitoring system and research infra- Working in karst environments without adversely affecting structure funding than that described for geochemical monitor- them is a major challenge in cave research and conservation. ing. It will be necessary to monitor 1) the local weather, hydrol- Is it possible to enact an international research policy regarding ogy, ecosystem and landscape above a monitoring cave system; cave research that scientists will follow? The varying factors and 2) in-situ cave measurements: cave meteorology and air of different local and federal restrictions on cave access and composition, water flow, water chemistry, calcite growth, and research permitting is an obstacle to be addressed. One as- calcite chemistry. pect of such a policy might be the coring and reconnaissance analysis of speleothems vs. the multiple whole sample removal Standardization, Archiving, and from caves. The karst community lacks an archival system for Conservation researchers such as that which exists for ice cores, yet speleothems are rarer. There are competing pressures of multiple Standardization of several practices in using speleothems for sampling trips needed if reconnaissance sampling is employed, climate reconstructions can help to yield not only better quality vs. the time and funding required for the recon approach. One datasets, but will help preserve cave environments for future way to enact such a policy, if it could be agreed upon, would be generations. Preliminary reconnaissance-type field missions to require a conservation plan as part of the “Broader Impacts” that employ minimally-invasive sampling can help to bet- section of a research grant proposal.

88 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Focus Group on Geochemistry and Climate

Geochronology understanding of the second set of processes will have direct application to important science questions regarding the opera- The road to high-fidelity, well-dated speleothem paleoclimate tion of modern karst systems. records is paved with more dates than at first appearance might be necessary, and more reproducibility than is often funded or CROSS-CUTTING THEMES WITH OTHER published. This will require laborious, high-resolution recon- DISCIPLINES structions of 20th century climate and, absent visible banding, identifying chemical, isotopic, or petrographic proxies that oc- Karst research has synergy with and bearing on other fields cur in annual cycles, and counting these annual cycles back of study including landscape ecology, microbial ecology; the from the present. The methodology for executing such records global carbon cycle; biogeochemical cycles; water resource exists or can be developed, but it will require significant ef- policy and decision support systems. Some of the links be- fort that can be driven by funding priorities. To improve the tween these fields and karst systems are discussed above, such accuracy of U-series age determinations, we will need an im- as in the section on the feedbacks between the global carbon proved understanding of initial 230Th/232Th values incorporated cycle and karst systems. Other links are posed by the following into speleothems. Studies of modern system, zero-age spele- questions that cross cut different disciplines represented in the othems, akin to studies of living corals to constrain marine ini- Karst Research Frontiers workshop: tial 230Th/232Th, will be valuable for addressing this issue. • What role do microbes play in calcite precipitation or dis- SUMMARY OF RESEARCH NEEDS solution and speleothem growth? What role do microbes play in incorporation of proxies in speleothem records In the next decade, the study of speleothems will have a large (e.g. C-isotopes)? impact in two areas of the earth sciences: (1) the understanding • What geochemical data and spatial/temporal resolution are of the major factors that have caused Earth’s climate to change needed for testing and advancing hydrologic modeling of in the past, and that likely will cause Earth’s climate to change karst systems? in the future, and (2) the understanding of those soil, vadose • How do changes in ecosystem functioning, both within zone, cave, aquifer and climatic processes that affect the iso- a cave and on the landscape above a cave, impact spele- topic and chemical composition of karst waters and the cave othem growth and incorporation of inorganic and biologi- deposits that precipitate from them. Important climate science cal proxies? goals will relate to the first point, whereas some significant un- • How can decision support systems be developed to maxi- derstanding of the second set of processes will be necessary mize benefits of karst resources to stakeholders while pre- in order to fully realize the first set of goals. The improved serving the natural capital of karst terrains?

Frontiers of Karst Research 89

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Focus Group on Caves and Karst as Model Systems in Geomicrobioligy

Focus Group on Caves and Karst as Model Systems in Geomicrobiology

Group Participants: Annette Summers Engel and Diana Northup, Leaders. Marcus Gary, Brett Gonzalez, Juan Gonzalez, Elena Hutchens, Dan Jones, Jennifer Macalady, John Spear, and Michael Spilde

INTRODUCTION environmental and civil engineers, and hydrogeologists) want to know how microbial processes alter the quality of water, the icrobial life evolved between 3.5 and 4.0 billion atmosphere, and the environment in general, as environmental years ago and evidence suggests that microbes have conditions may affect human health. played essential and ubiquitous roles in geological M Cave and karst settings are model systems where these research and ecological processes through time. In particular, most low temperature geochemical reactions are controlled, to a large goals can be easily and readily addressed by microbial scien- extent, by microbial metabolic processes. Through metabolic, tists. From a microbial science standpoint, caves and karst are enzymatic, or cellular catalysis, microbes participate in the pre- unique to study because of the link between the surface and the cipitation of minerals and influence rock dissolution, thereby subsurface (i.e. the so-called, “critical zone”), as well as the shaping and changing the earth’s landscapes through time. longevity of some karst settings. Based on in-depth discussions From an ecological standpoint, microbes serve as the energetic by Workshop participants, three main research themes and as- base of ecosystems, serving as a food source for higher trophic sociated questions emerged as priority future research areas. levels, and also producing ample, high quality energy through These research directions should yield promising practical (i.e. autotrophic metabolic pathways. economic) and intellectual results, as well as provide avenues for prosperous educational and outreach endeavors. Addition- Karst landscapes comprise 12.5% of the earth’s ice-free land ally, one key hope was that future investigations of microbial surface, coinciding with the global occurrence of carbonate sed- processes and activities in cave and karst settings should in- imentary rocks. Consequently, karst surfaces offer exception- volve multidisciplinary and international collaborations. The ally reactive interfaces for microbially induced and enhanced remainder of this report details the consensus of these discus- reactions occurring at air-rock-water interfaces. Considering sions and presents recommendations for cave and karst micro- the extent to which carbonate rocks comprise the rock record bial science future research directions. on earth and the depths to which karstification occurs in the subsurface, the microbial biomass within karst settings and at MAIN RESEARCH THEMES karst interfaces is potentially tremendous. Therefore, microbes and microbial processes are central to all of the karst-related Breakout session discussions centered on three main research sciences. themes, prompted by the State-of-the-Art presentations. Focus questions are presented with commentary, additional questions, In general, microbial scientists (e.g., microbial ecologists, mi- and future goals to focus the question. crobial taxonomists, microbiologists, astrobiologists, and pale- ontologists) aim to understand the diversity of microbial life on Biodiversity and Evolution earth, as well as how and when life evolved. Another goal for some microbial scientists (e.g., including geomicrobiologists, a) Who’s Home? All three domains of life (Eukarya, Bacteria, geobiologists, and geologists) is to be able to discern between bi- and Archaea) and viruses occur as microscopic life in caves and otic and abiotic processes in modern and ancient environments, karst. Among the Archaea and Bacteria, abundant novel organ- and the extent to which microbes have influenced geochemical isms representing new genera, families, and orders have been and geological processes, as well as how those processes have previously uncovered from caves and karst using culture-based affected the diversity of microbes. Because ~1.6 billion people and molecular phylogenetic studies. For many different cave depend upon the health of karst terrains and aquifers for their habitats, such as acidic cave-wall biofilms and surfaces, iron water supply world-wide (Williams, 1993), other types of mi- and manganese crusts on cave-walls, or microbial mats from crobial scientists (e.g., biogeochemists, organic geochemists, sulfidic cave streams, the question, “Who’s home?”, is getting

90 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Focus Group on Caves and Karst as Model Systems in Geomicrobiology

answered . Some caves are hotspots of microbial biodiversity, derstanding of microbial evolution and speciation hypotheses: with a prime example being associated with sulfidic caves and karst springs. Basic microbial biodiversity studies based on • How do microorganisms speciate in the subsurface? Is 16S rRNA gene surveys and phylogenies have shown that there gene transfer as active in the subsurface? are community and species parallels among caves on differ- • Could microorganisms identified from caves, and others ent continents. These habitats provide a model set of habitat in similar subsurface environments, have existed in this conditions for investigating biogeography and addressing the milieu for millions of years, slowly reproducing? question of whether “everything is everywhere; the environ- • Does the depositional setting of the original host rock in- ment selects.” fluence the modern microbial communities found in the subsurface? However, vast gaps still remain in our knowledge of microbial • Does allopatry influence the formation of new subsurface diversity in caves and karst, as most of the previous studies microbial species, especially in caves that are not well have been done in only a few caves. There is a fundamen- connected hydrologically to other voids? tal need to explore and document the biodiversity of different kinds of habitats in caves and karst systems. More baseline Basic research is needed in the whole of the microbial sciences data are needed to determine if microbes are endemic to caves with respect to mechanisms by which speciation can be due to or karst, or the subsurface in general, and if so, in what num- gene-swapping (including viral) interactions, and general eco- bers. Almost nothing is known about viruses in karst. The system services by viruses. need to know more about habitat biodiversity is not unique to caves and karst studies, but is perhaps more acute because of c) Do Microbes Adapt to the Subsurface Environment? the overall societal interest in understanding karst processes. Caves are often regarded as being very stable in temperature, Increasing our understanding of cave and karst biodiversity al- light conditions, and relative humidity. However, there are lows for better protection and conservation of these novel com- aspects of caves that can change rapidly, especially in hydro- munities. logically active systems (e.g., caves that flood seasonally or periodically). Ranges of physiochemical environments and b) What is the Source of Microbes in Caves and Karst? Al- habitats not only dictate the types of speleogenetic processes most a century ago, biospeleologists understood that microbes that might be prevalent, but also the different metabolic strate- were transported into cave systems via dripping water, air cur- gies that microorganisms might need given that particular set of rents, animals, and human visitors. Some cave microbes were parameters. In karst, these parameters could be related to the found to be merely a subset of microbial groups from the sur- speciation and availability of redox-sensitive elements, or the face that are known to be associated with overlying soil, includ- types and loading of carbon or other nutrient sources. Aquatic ing organic matter and detritus from plants, rhizosphere exu- habitat conditions will be distinct from subaerial conditions and dates, etc. From an ecosystem perspective, the metabolic types stresses, as microbial community composition and structure of allochthonous microbes can have a profound impact on the from subaqueous microbial mats are different than the compo- niche availability and the overall function of the ecosystem. sition and structure of communities living attached to cave-wall From a geological perspective, microbial life has a long history mineral surfaces. In more hydrologically stable areas (e.g., of existing and interacting in mineral environments and it may within the phreatic zone), there are no obvious “hot times” in be necessary to consider that the depositional history and geo- microbial community development. But, in caves like Cueva logical setting of the host rock in sourcing microbes in cave and de Villa Luz in Mexico, with large and erratic inputs of reduced karst. The discovery of chemolithoautotrophic communities in gases, the hypothesis would be that diversity may vary over caves has opened up a new suite of questions about the origin time. Does the ingress of reduced gases into cave environments of these microbial communities. Phylogenetic reconstruction bring with it new microbial biomass that originates from deep of 16S rRNA genes suggests that cave microorganisms from within the Earth’s plumbing? Lechuguilla Cave in New Mexico have marine ancestors. Most limestones, including the limestone in which Lechuguilla Cave Do surface-derived microorganisms die due to the oligotrophic formed, are marine in origin. nature of many caves, becoming food for higher level organisms, or can the surface-derived microbes function in the re- Several questions arose from discussions about the origin of stricted (i.e. based on physicochemical conditions) habitat microbes in caves and karst, and the implications of conducting and subsequently colonize caves? If so, then how do these research in cave and karst settings that could influence our un- microbes, once exposed to the harsh and erratic surface con-

Frontiers of Karst Research 91

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Focus Group on Caves and Karst as Model Systems in Geomicrobioligy

ditions, adapt to the more stable conditions of karst? Unlike cave ecosystems. The base of traditional food webs drawn for surface-derived transport in dripwaters, it is unclear the extent most cave ecosystems has centered on a large black box that to which air currents and ground waters bring microbes into the is referred to as “microorganisms”. Arrows come into and out cave environment. For this research to reach conclusions, the of this box, demonstrating fluxes and processes that involve geological and hydrological connectivity that exists, and has microbes. Inputs are generally considered physicochemical or existed, between surface and subsurface environments needs nutrient constituents important for microbial pathways. For to be better characterized. In the more stable, low-nutrient (i.e. photosynthetically-based ecosystems, a critical input is light low organic carbon) habitats, selection may lead to the stream- energy. Obviously, for cave ecosystems, this is not possible. lining of genomes and possibly to a reduction in genome size. Instead, inputs could be energy-rich, commonly redox-sensi- Fundamental questions need to be addressed about evolution- tive, compounds (e.g., methane, hydrogen sulfide, or other re- ary forces in the subsurface: duced substances such as iron or manganese), or allochthonous carbon compounds. Numbers are rarely added to these flux ar- • Are microbes adapting to subsurface conditions? rows, and future research needs not only to identify flux inputs • At what rate does adaptation proceed in the subsurface? and outputs to ecosystems, but to quantify flux magnitude. In • How do microbial mutation rates compare between sur- general, the role of allochthonous microbes to ecosystem func- face and subsurface environments? tion (e.g., heterotrophic productivity) has rarely been studied, • Which is faster: genetic change or transport of microbes? and only limited work has been done to characterize microbial role types within communities (e.g., autotroph or heterotroph), Investigation of these questions will be greatly aided by genom- or to understand the scale of microbial effects and controls, on ics and whole genome sequencing of communities of microor- ecosystem function. ganisms in the subsurface. Thus far, whole genome sequencing for cave microbes has not been done. However, we anticipate b) Are Nutrients Limiting for Microorganisms in the Sub- that the rapid advances in the field, from addressing the basic surface? Across different karst systems and habitats, geo- questions, “Who’s home?” to “Why are these microbes in this chemical constituents vary, which should lead to differences cave?” will be completed and will establish some evolutionary in microbial composition and roles played in the ecosystem hypotheses. Whole genomes of microbes from similar karst by microbial inhabitants. Future research investigations could habitats, possibly distributed throughout global systems, will fruitfully address the following questions that basically remain be relevant to fundamental questions about life processes in all unanswered: organisms (e.g. quiescence). • What roles do microorganisms play in biogeochemical cy- d) Does the Subsurface Habitat Influence Microbial Com- cling in the subsurface? Are nitrogen, phosphorus, or car- munity Composition? The geologic history provides a back- bon limiting for microorganisms in the subsurface? How drop against which we study microbial interactions with each do N, P, C, and S cycle in the subsurface? Are other trace other and the rock environment that influences them, but ac- elements limiting for microorganisms? tively metabolizing microbes also influence their habitat (e.g., • How do heavy metals cycle in the subsurface and do mi- through the production of acid by-products). One of the ques- croorganisms play a role in this cycling? Do microorgan- tions discussed was whether there is an indigenous subsurface isms scavenge rare earth elements? microbiota. If so, then surface-derived microbes would only • What is role of microbes in cycling these elements from supplement this subsurface community. To address this ques- the atmosphere, to the surface, to the subsurface through tion, in light of the aforementioned questions, we need to es- the critical zone? tablish whether there are similarities in microbial community composition among caves with similar habitat conditions. c) Are Microbial Community Structure and Function Linked? One of the outcomes of understanding the scales Ecosystem Function of microbial impacts to ecosystems is the characterization of niche dimensions, if microbes can be defined based on niche. a) How Are Microbes Central to Ecosystem Function in Microbial niche and ecosystem dimensions could extend from Different Types of Karst? Based on previous cave ecosys- the micron, to the meter, to the kilometer scale. From a con- tem research, microbes, as food sources or from autotrophy, servation perspective, the stability of these dimensions requires are key to the nutritional status and availability (i.e. quality characterization. This issue is not isolated to cave and karst and quantity of carbon substrates) and the energetic base of microbial science research, as the categorization of microbes

92 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Focus Group on Caves and Karst as Model Systems in Geomicrobiology

by niche is currently difficult to do in any field. them. The role that microorganisms play in decontaminating pollutants in the subsurface is not known either and the pos- New developments in genomics and proteomics are providing sibility exists that microorganisms may play a role in maintain- tools with which to assess microbial function in communities ing water quality in subsurface aquifers. and ecosystems. To investigate how community structure and function are linked we should apply these new techniques to cave and karst studies. PROSPECTS FOR THE FUTURE

Disturbance Justification

a) Are Microorganisms Agents or Victims of Disturbance? Microbial ecosystems are incredibly complex and difficult to Disturbance can be natural or may originate from human influ- study. Cave and karst microbial ecosystems provide model ence. The degree to which microorganisms act as agents of systems in which to conduct a range of scientific inquiries, as disturbance, or are influenced by natural or anthropogenic dis- detailed above. Not only are caves systems intrinsically inter- turbances, has received limited attention. Some excellent work esting and valuable, but they provide an environment with a has been done in Spain and France on the role of humans in reduced number of variables. Weathering is limited, or nearly causing damage to cave art through the introduction of carbon absent, in some cave systems. In oligotrophic systems, pre- and exotic microorganisms. Some work has been done on us- dation by macroscopic predators can be very limited. Inputs ing human indicator microbes, such as Staphylococcus aureus, and outputs are more easily defined. Thus, studies of complex E. coli, and high-temperature Bacillus spp. from overlying des- interactions can actually be studied in an environment with a ert soils, as monitoring tools to determine human impact on limited set of variables, making these ideal systems in which to caves. Beyond these few studies, however, much remains to study high-level microbial science questions. be studied to understand the role of microorganisms in environmental issues that relate to karst. Humans are increasingly Recommendations occupying karst landscapes, bringing pollution to karst aquifers and systems. The effects of this urbanization on karst microbial Within the microbial sciences, there has been concerted effort communities are unknown. Important future research direc- at the national (e.g., American Society of Microbiology) and tions and questions include: international (e.g., Society for General Microbiology, Interna- tional Committee on Systematic Bacteriology) level to make • What factors (biological and/or geochemical) cause distur- overall recommendations for future research needs in the field. bance to karst microbial communities? For cave and karst microbial sciences, we too assert that there • What constitutes disturbance to a microbe in the subsur- is a fundamental need for more funding, more investigators face? (including individuals and consortia), and more education and • Are microbes an aspect of disturbance in karst? communication among students within the cave and karst mi- • At what point does the enrichment of carbon threaten in- crobial sciences community. During breakout session discus- digenous microbial communities? sions, three general research needs were identified:

In oligotrophic karst systems, even low levels of carbon enrich- a) Studies that take place over varying spatial and tempo- ment may have an effect on subsurface microorganisms that ral scales: Varying scales, from mineral to mineral surface, are harmed by eutrophic conditions. Disasters, such as the pol- from cave to cave, and from continent to continent, will ad- lution of a cave by a gasoline spill, kill many organisms, in- dress questions related to what hydrogeochemical and ecologi- cluding microorganisms. Other microorganisms may be able to cal controls may be influencing microbial community diversity degrade contaminants and may move into the polluted area, re- and structure. Scalar investigations will also lead to a better un- placing native bacteria killed by the contaminants. Many show derstanding of endemism and biogeography issues. Moreover, caves and caves on private lands have septic tanks or sewage because it unclear how microbial strains within a described lines above them. When breaks or leaks occur in these systems, species evolve over time, changes in genomes through time carbon enrichment is extreme and many organisms associated can profoundly affect species designations. Therefore, more with the sewage are released into karst systems, harming the studies are needed to evaluate changes in microbial community native biota. There is a wide variety of contaminants that are compositions and structures over time in the laboratory and in now polluting karst systems and we know little about most of cave and karst settings.

Frontiers of Karst Research 93

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Focus Group on Caves and Karst as Model Systems in Geomicrobioligy

b) Better methods and tools for understanding presence needed. Currently, few caves or karst settings are logged con- or absence of microbes: Future culture-based or genetic in- tinuously because of instrumentation sensitivity and measure- vestigations cannot be exhaustive due to economic costs and ment drift, which often can lead to erroneous and nonsensical time commitments. Therefore, more inclusive, in situ, and interpretations. non-invasive techniques and tools to study microbial diversity and ecosystem function are needed. We expect that screening Sampling strategies that can acquire biomass without disturb- clone libraries and sequencing thousands of clones will not be ing sample integrity and community structure need to be de- feasible for large-scale projects. However, metagenomics and veloped. Techniques have been established for soft microbial proteomics hold a great deal of promise for addressing these mats, but sampling corroded rock, sediments, or solid materials questions. will require new strategies. Maintaining the integrity of cores into punky rock is extremely difficult and requires the develop- c) Good model lab for studying complex processes: A clear ment of new light-weight, more effective drilling technology consensus was reached that a model cave or karst system is that can be employed in an aseptic manner. needed in order to address all of the microbial science questions, from an ecological perspective to the geological perspec- Instrumentation and logging needs could be met with financial tive. A model system could allow not only for basic observa- and technical investments in the development of autonomic mi- tional and discovery-based research, but also experimentation crobots. Microbots could be essentially remote sensing tools and manipulation of the system to test disturbance issues. “on a string” and could be used in deep karst aquifer wells or in spring fissures. Microbots could be useful to other scien- Technical Barriers tific disciplines, such as hydrology, whereby 3-D mapping and characterization of habitats could be done. Microbots with Research in caves and karst is difficult to conduct. These are geochemical sensing and sampling capabilities would be ex- humid (~97-100% relative humidity) and potentially corrosive tremely useful in some karst environments. Partnerships with habitats. Transport of equipment needed for research studies astrobiologists may facilitate this development. into caves requires hauling and maneuvering over obstacles, and sometimes through narrow passages; this can easily dam- Cave and karst science has notoriously been perceived as age sensitive instrumentation. Currently, some scientists have “cheap”. Despite the fact that molecular techniques in the designed instruments to work in the cave habitat, but this has microbial sciences are advancing quickly, and some cave and been a costly and time-consuming endeavor. Future efforts karst researchers have already started to take advantage of the should address the need for stable, durable, economical instru- vast array of methods, including pyrosequencing and whole mentation and equipment. Additionally, future work should genome studies, most of the cave and karst microbial scien- also work toward standardization of scientific practices among tists are unfortunately lagging behind the technology cutting- investigative teams. edge. These researchers are limited by budgetary restrictions; molecular methods are expensive, from the standpoint of the Required tools necessary instrumentation and training, plus long-term financial commitments to keep the instruments running. Much like Characterizing, monitoring, and analyzing microbial communi- other molecular microbial scientists, cave and karst scientists ties in situ will need to be addressed. Given current limitations need cheaper sequencing methods, such as retrieving >1000 bp (e.g., financial, technical, personnel, and physical), more du- for pyrosequencing, or faster and more reliable methods be- rable and stable, yet smaller and cheaper, tools are required to yond 454 sequencing and we need sequencing from smaller perform microbial science experiments and to conduct in situ amounts of DNA as biomass is limited in some critical study studies in cave and karst settings. areas. Similar to other types of investigations in the Microbial Sciences, computational horsepower, with better annotation, Instrumentation that can detect microbial metabolic activities, phylogenetic, and statistical tools, are needed. such as gas production or consumption (e.g., respiration), will be useful to understanding microbial ecosystem function. Es- Standard Practices sentially, a gas chromatograph interfaced with a mass spec- trometer would be ideal, and should be a tool developed in the The application of traditional microbiological and molecular future for cave and karst research. Moreover, better, stable, methods may not be appropriate for cave and karst investi- quick-responding, and robust, data-logging capabilities are also gations. Traditional microbiological culturing methods will

94 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Focus Group on Caves and Karst as Model Systems in Geomicrobiology

not work well in caves, particularly for microbes from olig- studies, involving them in field and laboratory work. Science otrophic habitats. However, cultivation studies are needed to fair projects are one of the tried-and-true methods to stimulate study the physiology of indigenous organisms, so better cul- student involvement, but summer programs are also fun and tivation methods need to be developed. Several researchers effective. We have also begun initiatives to bring public school have asserted that in situ inoculation and incubation are critical teachers into cave and karst research so that they become part of to successfully capturing the biodiversity of organisms from our recruiting team. Working with existing national programs, caves sites. Rigorous tests of this hypothesis are necessary to we need to expand these efforts and improve our ability to sell determine whether culturing efforts of cave samples are grow- karst systems as model laboratory for studying fundamental ing the weeds or the indigenous microbial community. Other processes that occur on the surface and subsurface. Money to technical aspects that can affect the quality of the results of support such initiatives is always scarce and it’s important to karst geomicrobiological studies include methods of storage expand our efforts and apply our creativity to obtain new fund- and transportation of collected samples and DNA extraction ing. and amplification. Cave samples can represent significant challenges in DNA isolation, purification, and amplification. Es- International Cooperation tablishing suggested best practices for the technical aspects of geomicrobiological studies in caves would be useful to both A recurring theme of the Workshop was the need to enhance established researchers and researchers from other fields enter- the access to cave and karst information, to encourage linkages ing into karst research. and communication among karst scientists globally, and to establish collaborative digital work spaces that lead to knowledge Within the broad field of Microbial Sciences, there are obvi- discovery from existing karst data. New initiatives by the Uni- ously a number of different ways that a scientist could conduct versity of South Florida, the University of New Mexico, the Na- microbiological and geomicrobiology research. There have tional Cave and Karst Research Institute, and the newly emerg- been debates in the past regarding best practices for differenti- ing International Cave and Karst Research Institutes Network ating abiotic from biotic processes, which is a non-trivial issue. (ICKRIN) are seeking to establish a Karst Information Portal Resolving best practices in cave and karst research, while not (KIP) that will facilitate many of needed linkages and access to glamorous, affects all the core studies that we wish to carry out electronic version of various karst resources. KIP will launch in cave and karst geomicrobiology, and can help researchers in June of 2007 and efforts will intensify to add to the content new to the cave environment understand the differences and and communication channels. A pilot project that is part of KIP challenges that caves and karst offer. Further research in this will provide a test bed of scanning electron micrographs in an area can advance our science and make important contributions institutional repository. An accompanying collaborative work to the general field of geomicrobiology. We should address best space for discussing morphological features in the images will methods for preparation of samples for electron microscopy be provided through a Drupal/Gallery installation. in order to accurately visualize samples without preparation artifacts or contamination. We need to encourage multidisci- Despite these promising ventures, research and pilot projects plinary studies that bring together geoscientists and bioscien- are needed to investigate: tists in research efforts to produce the best science. • What methods best facilitate access to karst information, Education and Outreach especially the grey literature and imagery? • How do we effectively deal with intellectual property is- The “graying” of the karst community is a significant concern. sues? New educational initiatives and enhancement of existing cave • How can we promote digital collaboration and discussion and karst programs are needed. To attract fresh blood to the to promote knowledge discovery? field, we need to expand and enhance educational efforts that target middle and high school students who find caves and re- These collaborative efforts, using digital technology, will en- search in caves fascinating. Some karst scientists are initiating hance knowledge discovery in cave and karst science. programs to bring these students into cave and karst research

Frontiers of Karst Research 95

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Focus Group on Ecosystem Function

Focus Group on Ecosystem Function

Group Participants: Kevin Simon, leader. Daniel Fong, Lara Hinderstein, Bridget Maloney, Robert Payn, Michael Vernarsky, Frank Wilhelm

MAJOR ISSUES processes that transform that energy and ii) identifying the key factors regulating the spatial and temporal variation in karst. he ecosystem function working group identified two In particular, identifying the presence of “hotspots” and “hot major topics for future research: times” for ecosystem function and the factors that lead to those situations should be addressed. Areas and times of enhanced T connectivity between the surface and subsurface are likely to What Limits Productivity in Karst? be important in this issue. Once the patterns and drivers of variation in ecosystem function are better understood, empha- It is now known that microbes are key mediators of energy flux sis should shift to the consequences of spatial and temporal in food webs of karst systems that are exclusively heterotrophic variability for karst ecosystems and the other ecosystems con- as well as those fueled by chemoautotrophy. Thus, explora- nected to karst groundwater. tion of the factors that are likely to limit microbial productivity should be of primary concern. In exclusively heterotrophic JUSTIFICATION systems, dissolved organic matter (DOM) has been implicated as an important energy source for microbes and study of the Broad Perspective influence of both the amount and composition of DOMon microbial productivity is required. The importance of other Groundwater and karst systems in general have been consid- types of organic matter, such as leaves and wood, should not ered to be organic carbon limited. In many ways, karst basins be ignored considering the data available to date about energy are model systems for exploring organic carbon limitation in sources used by cave food animals are somewhat limited. A groundwater and the role of dissolved organic carbon and mi- few data suggest inorganic nutrients (nitrogen and phosphorus) crobial food webs in general for several reasons. First, the pres- do not limit microbial productivity in heterotrophic systems, ence of caves allows unparalleled access to aquifers allowing but this needs further exploration. The role of inorganic nu- a truly 3 dimensional approach to study surface-subsurface in- trients in regulating productivity in chemoautotrophic systems teraction. Second, the exclusively heterotrophic food webs and should be a productive avenue of research. There are some data availability of systems with differing types of organic carbon that suggest microbial productivity may be limited because of input allow study of detritus in fueling food webs without the animal grazing. This and other potential limiting factors (e.g., confounding effects of photoautotrophy. Indeed karst contains flooding, temperature) need to be explored. Virtually no data model systems for understanding microbial-based food webs. are available regarding energy limitation and flux in terrestrial food webs and there is a clear need for research addressing pro- In karst, the drainage network has essentially been moved from ductivity in terrestrial habitats and their connectivity and rela- the surface to the subsurface. As a result, processes in the sub- tionship to aquatic habitats. surface and ground water have replaced the ecosystem services typically provided by surface streams and rivers. Understand- Spatial and Temporal Variation in Ecosystem ing subsurface processes should increase our understanding of Function. the ecosystem services provided by karst ground water. For example, the transport of some pollutants is tied to the organic Evidence is mounting that ecosystem function in karst varies matter to which they sorb. As a result, organic matter process- spatially and temporally. The key to this variability is likely ing in karst landscapes may play a key role in the transport and hydrology because of its role in linking subsystems within transformation of pollutants in ground water. karst basins as well as linking the subsurface to other systems (e.g., surface soils and vegetation, and marine systems). Future Karst Perspective research needs to focus on two primary issues: i) quantifying the spatial and temporal patterns in the delivery of energy and The ecology and evolution of animals in karst has been strongly linked to energy/productivity. Understanding what drives the

96 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Focus Group on Ecosystem Function

productivity of karst food webs will have large implications for There are clear transdisciplinary intersections between ecosys- our understanding of the evolution of cave animals and other tem ecologists and other scientists, for example hydrologists related issues such as patterns of biodiversity in karst. Karst (quantifying water and solute flux through epikarst). Linking ecosystems are highly susceptible to impacts from anthropo- spatially explicit models of hydrologic flux to energy and nutri- genic activities and issues such as land use and climate change ent flux in karst is needed. Thus, close collaboration, including will influence the spatial/temporal pattern of water flux and joint manipulative experiments, must occur to advance the field surface vegetation/soils in karst. Understanding and predicting of ecosystem science in karst. The availability of karst basins the ecological consequences of such changes in karst is para- (and/or portions of basins) in which controlled, long-term ma- mount to protect and manage these ecosystems for their contin- nipulation and modeling can be completed will facilitate such ued existence. For example, understanding the hot spots and collaboration. Ideally, these experimental basins would have hot times of ecosystem function, and factors that drive them, good access to the subsurface (i.e., cave passage) to allow de- should guide human activity on the surface. tailed sampling within karst. The ability to instrument and ma- nipulate the surface (e.g., change surface vegetation, soils, and APPROACHES water flux) would be highly desirable.

Comparative and descriptive research is still needed to under- The lack of basic data about ecosystem function (e.g., standing stand food web structure and quantify the spatial and temporal stocks and fluxes of organic matter) and the development of patterns of energy flux in karst. However, we strongly empha- quantitative methods to sample the epikarst and other less ac- size that future studies of ecosystem function include manipu- cessible locations also hinders hypothesis generation and test- lative experiments. Such manipulative experiments should being in karst. gin at small spatial scales, but eventually be expanded to large scales to test hypotheses related to issues such as connectivity TOOLS within subsystems in karst and between karst and other ecosystems. The use of comparative studies (e.g., gradient stud- The primary tool needed to advance ecosystem science in karst ies) and modeling should also be pursued. Most research con- is a set of karst ecosystems that can be extensively monitored ducted to date has been of limited duration and we believe a and manipulated. Ideally, the size of such systems should range “LTER” approach for long term data collection and integration from small subsystems to entire karst drainage basins. Small is needed. subbasins or caves would be sufficient to test a variety of hypotheses and generate data. Ideally, at least one karst basin MAJOR BARRIERS would be fully instrumented (precipitation, water flux) and available for surface and subsurface manipulations. Such a lo- Two major barriers to advancing ecosystem science in karst cation would be an ideal site to foster collaborations between are: i) the lack of collaboration between biologists, hydrolo- ecosystem ecologists, hydrologists, geochemists and microbi- gists, geochemists and microbiologists; and ii) the lack of karst ologists. basins which can be monitored and manipulated for long periods.

Frontiers of Karst Research 97

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Focus Group on Subterranean Biodiversity

Focus Group on Subterranean Biodiversity

Group Participants: David C. Culver, leader. Penelope Boston, Mary Christman, Valerie Collins, James Godwin, Horton Hobbs, Brian Holmes, John R. Holsinger, Thomas Iliffe, Jean Krejca, Jerry Lewis, Kathleen O’Conner, Tanja Pipan, Katie Schneider, Steven Taylor, and Maja Zagmajster

INTRODUCTION Many species have relatively extensive ranges, up to hundreds of kilometers, even when cryptic species (genetically distinct ubterranean biodiversity not only encompasses the di- but morphologically identical) are taken into account. It also versity of subterranean life, it also encompasses a rich misses the point about cryptic species. Why are they so nu- diversity of research interests and research directions. merous in caves? At a superficial level the answer is obvious. S Different caves have strikingly similar environments, and ad- The bread and butter of subterranean biodiversity studies is the description of new species, of which there is a seemingly aptation should result in convergent morphology. However, the never-ending supply. It seems to be never-ending because of dynamics of this are not at all clear. There is often a mismatch the high levels of endemism of the obligate cave fauna. More between genetic convergence and morphological convergence. intensive collections and collections from previously uncollected caves typically yield new species. Without a continuing Stygobionts and troglobionts with relatively large ranges, or at alpha-taxonomic enterprise, further analysis and work is diffi- least groups of cryptic species, are often of special interest to cult if not impossible. This problem is neither new nor unique students of evolution. It is important to note that the prime to subterranean biology, but it is perhaps especially acute in candidates for model systems for the study of evolution and subterranean biology. New models for species description and development are among those species where the line between species identification need to be explored. Among the sugges- species is blurred. Examples include different populations of tions that came out of the workshop were: the Mexican cave characin, Astyanax fasciatus, in the Sierra de El Abra and different populations of the amphipod Gam- 1. Training of “para-taxonomists” who can identify de- marus minus in West Virginia caves. One of the most striking scribed species, and recognize undescribed species. cases is that of cave populations of Asellus aquaticus in Slove- 2. Training of taxonomists who are not group specialists but nian caves. It appears that this species has invaded the same rather subterranean specialists. The emphasis here is on cave system (Postojna-Planina Cave System) at least twice, and species description rather than high level systematics. probably three times! 3. Increasing use of foreign taxonomists. What has not been done is the simultaneous mapping of genetic Two research themes emerged from the workshop. One was and morphological diversity of any of these “problem” species the problem of “cryptic” species that occur in different caves on a geographic scale. Do molecular genetic (i.e., barcoding) and cave regions, and the other was the problem of bias-free differences correlate with morphological differences? Are ge- mapping of subterranean biodiversity on a regional and conti- netic and morphological variations at a site correlated? Given nental scale. the frequency of cryptic species in caves, they are good model systems for the study of this phenomenon in general. CRYPTIC SPECIES MAPPING SUBTERRANEAN Most cave-limited species (aquatic stygobionts and terrestrial BIODIVERSITY troglobionts) have highly restricted ranges, often limited to a single cave or small group of nearby caves. No doubt the Our capacity to store and display locational information has in- standard view of most evolutionary biologists is that cave ani- creased exponentially over the past decade. Several large data mals invade or somehow become isolated in caves, speciate, bases of information about stygobionts and troglobionts exist and disperse little if at all. In fact, this is a rather uninteresting in both the U.S. and Europe, and we know at least the broad scenario, one that holds little interest to biologists who do not outlines of patterns of subterranean biodiversity in selected study caves. What it misses is that it is not true in general. caves throughout the world. Several attempts at a quantitative synthesis have been initiated but they have been constrained by

98 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Focus Group on Subterranean Biodiversity

several factors: ing general cave fauna inventories. Although to our knowledge no study of sampling completeness using pitfall traps exists, 1. Variation in intensity of sampling. Intensity and frequen- it should be relatively easy to determine sampling sufficiency cy of sampling in different caves varies, usually in an un- using a large number of unbaited traps which are repeatedly known way. sampled. At each pitfall trap site temperature, relative humid- 2. Sampling bias. Records of species from a cave does not ity, soil category, and organic carbon would be measured. necessarily mean all cave-limited species were sampled. 3. Sampling incompleteness. New species are being discov- Finally, a standard 1 g of cave soil could be sampled by each ered even at well sampled caves, such as Postojna Cave in pitfall trap site. They would be covered with sucrose lysis Slovenia and Vjetrenica Cave in Bosnia & Hercegovina. buffer to break open cells and stabilize the DNA. To assess and compare community microbial diversity, DNA will be ex- There was a consensus that a broad scale geographic analysis tracted and purified to provide the basis for amplifying a 500 of subterranean diversity was a very worthwhile goal, one that bp region of the genome for running on a gel using denaturing would yield insights into the control of subterranean diversity, gradient gel electrophoresis (DGGE). This will provide a com- assuming that sampling problems could be solved and that the munity fingerprint. As each band roughly corresponds to one appropriate environmental parameters could be measured. species and the brightness of the band corresponds to numerical dominance, we can assess species richness and relative abun- The approach to the sampling problem was not to emphasize dance across sites. Bands can be "picked" with pipette tips completeness, but rather to emphasize quantitative compara- and sequenced to give phylogenetic information about species bility. In order to accomplish this, three surrogates for total present. diversity were proposed: A total of 250 caves in North America and 250 caves in Europe 1. Epikarst copepods collected from ceiling drips as a sur- should provide an adequate sample. In each continent, between rogate for total stygobiotic diversity 10 and 12 regions would be sampled. In North America, these 2. Troglobionts collected in pitfall traps as a surrogate for would include: total troglobiotic diversity 3. Microbial community fingerprinting as a surrogate for to- • Florida lime sinks tal microbial diversity. • Appalachians • Interior Low Plateau Quantitative sampling of epikarst copepods was pioneered by • Driftless Area Dr. Tanja Pipan at Karst Research Institute in Postojna, Slo- • Ozarks venia. Her technique, involving continuous filtering of drip • Black Hills water, is particularly appealing because it involves long term • Edwards Aquifer/Balcones Escarpment sampling (months) and because she has demonstrated the con- • Guadalupe Mountains ditions for sampling completeness in terms of number of sam- • Scattered western areas ples and length of sampling. Epikarst species diversity if high, sometimes even surpasses the rest of the stygofauna in a cave. For each cave, location, altititude, cave length, depth, entrance At each drip, temperature, pH, dissolved oxygen, dissolved or- size and aspect, and number of entrances would be recorded. ganic carbon, conductivity, ceiling thickness, drip type (from For each region, land use, land cover, and rainfall would be formation or from ceiling), ceiling thickness, and fecal coli- recorded. These data should allow a predictive model of cave form would be measured. biodiversity.

The most standard technique for sampling terrestrial cave fau- The big questions are what are the patterns and what is the role na, aside from visual inspection, is the use of baited or unbaited of resource availability, especially organic carbon, in determin- pitfall traps. Such traps are used by almost all biologists doing these patterns.

Frontiers of Karst Research 99

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Understanding the Tempo and Mode of Evolution

Understanding the Tempo and Mode of Evolution: Cave Adaptation as a Model System

Group Participants: Megan Porter, Leader. Katharina Dittmar, Ben Hutchins, William Jeffery, Tristan Lefébure, Pierre Paquin and Meredith Protas

lthough cave-adapted animals have been studied for in the cave environment that have not yet been identified. over a century, only recently have they emerged as Apromising model systems for understanding the tempo and mode of evolution. With advances in molecular and genomic techniques, including the increasing accessibility and affordability of full genome sequencing, it is now possible to take advantage of these model systems to understand general questions in evolutionary biology that will have interdisciplinary impacts. The unique suite of characters associated with the cave-form, termed troglomorphy, is fascinating and significant due to the convergent nature of both constructive and regressive traits across diverse taxonomic groups, which provide the opportunity to understand the specific roles of mutation, selection, development, and gene network interactions in trait evolution. Moreover, many of the traits associated with cave adaptation – increased lifespan, increased obesity, modes of locomotion, multi-modal sensory adaptation, and the loss of vision – are of Figure 7. Examples of proposed cave-adapted model spe- critical importance from a medical perspective. The following cies, and where possible examples of the surface form of the main research themes focus on our assertion that cave species same species, exhibiting troglomorphic convergence. Asellus aquaticus cave form (A) and surface form (B); Astyanax mexi- should serve as model systems to address important questions canus cave form (C) and surface form (D); Gammarus minus (E); in evolutionary biology. undescribed species of stygobiont dytiscid beetle (F); and a nymph of the planthopper Oliarus polyphemus (G). Photos re- THE CAVE FORM produced with permission from M. Protas (A,B), W. Jeffery (C, D, G), D. Fong (E), and K. Miller (F).

Cave adaptation is characterized by distinct morphological, physiological, and behavioral attributes (Fig. 7). The most MODEL SYSTEMS commonly cited, and the most obvious, characters are the loss of ocular structures and pigmentation. Compared to their sur- Model systems are crucial to evolutionary biology studies, es- face counterparts, some cave-adapted animals are physiologi- pecially where multiple, cross-disciplinary avenues of research cally characterized by lower metabolic rates, increased life focus on a single species. This approach creates a communi- spans, and changes in reproduction, while behaviorally some ty of scientists contributing knowledge on a few of the most cave species exhibit less activity and aggression. Despite pre- promising systems, builds a strong informational foundation vious and ongoing investigations describing cave adaptation, on which to conduct research, and generates momentum for few studies have looked at these physiological and behavioral funding and scientific progress. Major advances in our under- characteristics comparatively across subterranean species. Re- standing of the development and evolution of diverse organ- cent research is finding that not all of these typical traits may isms and life processes have come from studies of Drosophila be present in all cave-adapted species. On the other hand, there melanogaster (fruit fly), Caenorhabditis elegans (nematode), may be traits, such as fat deposition, that are common in cave- and Danio rerio (zebrafish). adaptation that have not yet been well studied. These findings illustrate that studies of the typical ‘cave-form’ are still required, In contrast to these model organisms, cave-adapted species as there may be traits that are not universal to the cave environ- present unique features rendering them more suitable to serve ment, or there may be crucial traits that are a result of evolution as model systems in evolutionary biology (Fig. 7). The cave-

100 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Understanding the Tempo and Mode of Evolution

form has a suite of characters where trait polarity can be deter- Vertebrata mined by comparison with surface relatives (if existing), making trait loss easy to identify. Characidae, Characiformes, Teleostei, Astyanax mexicanus (De Filippi 1853) Among the cave species currently known, there are many found Currently, the only subterranean model system in evo- across a range of phylogenetic distances (populations within lutionary biology, although relatively under utilized, is Astya- one species to species spanning different phyla) that have ac- nax mexicanus, the Mexican cave tetra. This species meets all quired the cave-adapted traits independently (i.e. convergence of the listed criteria (above), and has been the focus of evo- of the troglomorphic form). Strikingly, and rare throughout lutionary and developmental studies since its discovery in the evolution, similar cave-adapted phenotypes can be found from 1930s. As such, A. mexicanus is the best candidate for genome invertebrates to vertebrates. This characteristic allows for com- sequencing, which will allow for large-scale investigations of parisons of the evolutionary mechanisms of particular traits genotype-phenotype linkages leading to the observed troglo- across different taxonomic groups across a wide range of ge- morphic forms. However, this will not be a simple undertaking, netic distances. Moreover, the primary environmental cue driv- because in order for the most interesting evolutionary analyses ing the evolution of cave-adapted traits – the perpetual absence to be performed, this requires having the unique ability to ob- of light is obvious, and contributing environmental conditions tain sequences from a representative of both the surface and are relatively straightforward to characterize (e.g., temperature, cave-adapted form. Much like the sequencing efforts for other nutrient sources, water stress, oxygen concentration, etc.). As model organisms, genome sequencing of A. mexicanus will ne- both subterranean terrestrial and aquatic species have cave- cessitate a multi-institutional, and likely an international, ef- adapted forms, evolutionary and developmental comparisons fort. across these habitats can help to understand what environmental parameters, in addition to the absence of light, are strong Arthropoda drivers of cave adaptation. For these reasons, caves contain unique organismal systems for evolutionary studies that should Crustacea serve as model species but have been under utilized. Asellidae, Isopoda Asellus aquaticus (Linnaeus 1758) Lastly, we suggest that cave-adapted species serving as model Similar to A. mexicanus, this isopod species contains systems must have three important characteristics: 1) the abil- both cave-adapted and surface populations that are distributed ity to be bred and raised in laboratory settings; 2) the ability to throughout the Dinaric Karst region and attempts to rear indi- produce large numbers of fertile offspring; and 3) the existence viduals in laboratory settings are underway. While a number of of closely related surface form(s), either in different popula- phylogenetic and phylogeographic studies have been done in tions of the same species or among a group of closely related this group, investigations of trait evolution have not yet begun. species. These criteria are important in establishing a useful Assuming A. mexicanus genome sequencing is accomplished laboratory organism and will help protect cave species that oc- in the near future, genome sequencing in A. aquaticus, of both cur in a single location or are found in low abundances. a cave and surface form, would allow for unprecedented comparisons of trait evolution among vertebrate and arthropod It would be ideal to establish model organismal systems from models. the main taxonomic groups found in subterranean systems, representing both terrestrial and aquatic species. However, for this Gammaridae, Amphipoda to be feasible, a much greater knowledge of taxonomic rela- Gammarus minus (Say 1818) tionships and ecologies, including abundance and geographic In terms of trait evolution and selection, the amphipod distributions, are needed. The following list represents our Gammarus minus represents the best-studied arthropod species. suggestions of species where research efforts should be focused Furthermore, G. minus can be bred in laboratory settings with to evaluate their potential to serve the entire evolutionary biol- minimal care. Unlike many cave species that show complete ogy community as model systems. These species were chosen absence of eyes and pigmentation, different populations of G. because a base level of knowledge already exists for evaluating minus exhibit varying degrees of eye and pigmentation reduc- their potential as model systems and upon which future studies tion in populations that have independently colonized subter- can build. Other species may be suitable, but require much ranean habitats. This range of reduction allows for the inves- more basic ecological investigation before evaluation is pos- tigation of different developmental and genetic mechanisms sible. leading to trait loss within populations of a single species.

Frontiers of Karst Research 101

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Understanding the Tempo and Mode of Evolution

Hexapoda and adaptation. Caves offer a unique system in which adapta- Cixiidae, Hemiptera, Insecta tion to a particular environment drives the evolution of specific Oliarus spp. (Plant hoppers) phenotypes where both the environment and resulting traits Most described troglomorphic species occur in lime- are known with certainty. Because caves are found around the stone caves; however, not all subterranean habitats are formed world, the cave environment is a repeated habitat where cave- in limestone. Other cave types include lava tubes, which are adapted phenotypes occur independently in geographically iso- young relative to most limestone caves, yet still contain cave- lated locations, as well as across phylogenetically diverse taxa. adapted faunas with reduced eyes and pigmentation. The rela- The following research themes illustrate how cave species can tively young age of lava tube habitats make troglomorphic lava serve as exceptional model systems for answering broad evo- tube species interesting for investigating which adaptive traits lutionary questions. appear first. Among these taxa, the Hawaiian planthopper genus Oliarus consists of many related species and subspecies How Can We Understand the Relationship adapted to lava tubes, which are closely related to species from Between Convergence and Divergence? another surface dwelling genus. This species group offers a promising comparative system for studying the convergence During evolution, there is a trend towards divergence, and much of older, limestone occurring, lineages and much younger lava of evolutionary study is devoted to understanding the resulting tube species. animal diversity. The field of ‘evo-devo’ – the study of the evolution of form – is aimed at understanding how developmental Dytiscidae, Coleoptera, Insecta genes and patterning mechanisms result in the origins of new Stygobiont diving beetles structures and diverse body plans. However, this focus on di- Two subfamilies of diving beetles – Hydroporinae versification represents only a piece of the evolutionary puzzle. and Copelatinae – contain closely related species adapted to Understanding how phylogenetically disparate taxa look simi- the subterranean (groundwater) and epigean realm. To date, lar is a complimentary line of research that receives much less most research of these species has concentrated on phyloge- attention. Convergence is a widespread evolutionary phenom- netic and taxonomic questions. Apart from the usual cave as- enon. Similar modifications and reductions of ocular structures sociated convergent traits, all cave forms are micropterous (e.g. has been observed in species adapted to a number of different exhibit reduced wings), and would represent an excellent study light-limited habitats in addition to caves, such as deep-sea system for the evolution of wing reduction and loss, and its dwellers, parasites, edaphobites, and burrowing animals. Yet, consequences for locomotion. the interactions between convergent and divergent phenotypes and genotypes and how these interactions affect morphologies Eubacteria have not been comprehensively investigated. Focusing research on particularly obvious convergent morphological traits Although there are many examples of macro-organisms adapt- (e.g. depigmentation) and the evolution of associated homolo- ed to cave life, only recently has attention be given to the possi- gous genes across taxonomies will help our understanding of bility of microbial subterranean adaptation (see report by Focus general trait evolution, as well as the connection between se- Group on geomicrobiology). It is still unclear what microbial quence evolution and protein structural constraints. adaptations may be driven by the cave environment, or subterranean habitat in general, and whether or not these are similar Cave-adapted species are highly convergent at the morphologi- to those adaptations observed in macro-organisms (e.g., de- cal level, even across disparate taxonomic groups. This con- creased growth rates or metabolic rates). Identification of cave- vergence has led to the existence of rampant cryptic genetic adapted microbial species will provide model systems where diversification in classically determined morphologic species potentially shorter generation times and smaller genome sizes (See Biodiversity section). The common occurrence of this will allow for manipulation studies and where mutation rates discord between morphology and genetics in cave species of- can be measured directly. fers a system that is ideal for developing methods that integrate genetic information into classical taxonomic methods. More- EVOLUTION IN KARST: MAIN RESEARCH over, because convergent traits can be investigated across a THEMES range of genetic distances – from morphologically dissimilar populations with similar genetic backgrounds through similar One of the overarching themes of evolutionary biology is to un- traits across different phyla – this system allows for investiga- derstand the tempo and mode of evolution, including selection tions of how convergence is achieved and how the evolution of

102 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Understanding the Tempo and Mode of Evolution

convergence and divergence are related. In particular, genomic comparisons to determine how much of the cave form is geneti- approaches will allow biologists to investigate the relationship cally hard-wired versus environmentally driven. Relative to of convergence at multiple levels of biological organization selection and adaptation, what aspects of the environment are (morphology, molecular, genomic, genome networks) and to most important? The typical cave environment is thought to determine how much of the genome is devoted to diversifica- be nutrient limited. Although few studies have quantitatively tion versus convergence. studied this feature, nutrient limitation may be a major factor influencing many of the typical cave-associated features, such Convergent structures are most often assumed to be the result as increased life span and reduced metabolic rates. The ab- of direct selection for a particular trait. However, work in A. sence of light may also be a strong factor, as eye reductions mexicanus suggests that eye loss may be the result of pleiot- and loss are common in a wide variety of environments that are ropy from selection for constructive traits related to feeding light limited, such as deep soil and marine habitats. structures. This leads to questions about how much of the cave form is truly adaptive, how much are the adaptive structures What is the Timescale of Evolutionary driving the rest of the observed morphology, and are there any Change? traits the sole product of genetic drift? This also brings into focus the problem of developmental constraint, which is a ma- The karst environment contains a preservation of information jor tenant in the field of evolutionary developmental biology. that is available in few surface habitats where weathering of- Can constraints channel the evolution of cave-adapted traits? ten removes the timing record of past geologic or geomorphic Although there has always been argument about the importance events. Karst, however, is partially removed from weathering of selection versus drift as drivers of evolution (selectionist- processes, and therefore contains an environmental depth with neutralist debate), the focus now needs to shift to the relative respect to recording past geologic and climatic events. Essen- roles of each. Cave-adapted species present an opportunity to tially, karst provides windows into the evolutionary past, where comparatively investigate the mechanisms and evolutionary time scales can be measured across generational, ecological, forces responsible for convergence, and to determine whether evolutionary, and geologic time. Studying the generational these are similar across disparate taxa. time scale is critical to understanding the evolution of the cave form, including questions such as how long are generation To What Extent is the Evolution of the Cave times in cave animals relative to surface relatives? Generation Form Controlled by the Nature of the Karst time affects the fixation of mutations, and how rapidly selec- Environment Versus Intrinsic Evolutionary tion can act on advantageous alleles, leading to the convergent Mechanisms? cave form. Ecologically, caves contain simplified ecosystems in highly stable environments where it may be possible to mea- Diversification, and ultimately speciation, is often the result of sure the rates of incoming invaders and the rates of population individuals successfully colonizing new habitats. However, and species extinctions. When considering evolutionary times- what makes a particular species capable of being a success- cales, caves are well constrained geologic environments con- ful colonizer? Pre-existing genetic variability? The evolution taining an excellent record of time from multiple sources; dates of novelty? Once successful colonization has occurred, what can be determined for the age of the rock, the clastic sediments, are the strongest environmental parameters driving evolution and speleothems within a particular cave system, all providing of particular traits? Are there any traits that are a phenotypic information on the age and stability of the environment through response to environmental constraints rather than the result of time. changes at the genetic level? How much of a role do phylogenetic constraints play in trait evolution (e.g. does your evolu- In many phylogeographic studies, ages are estimated for a par- tionary history matter)? Are there a limited number of ways to ticular lineage-splitting event and then these ages are compared lose traits regardless of the environmental driver, or does the to the timing of known geologic or climatic events a priori to environment drive the manner in which traits are lost? Karst search for possible causes. Well-studied karst systems allow settings offer repeated environments where these types of ques- for true hypothesis testing, where hypotheses about subterra- tions can be investigated. Obviously, the subterranean habitat nean colonization, lineage diversification, speciation, and trait exerts strong evolutionary pressures on inhabitants, as evi- evolution can be generated based on the geology of the system denced by the particular suite of traits characterized by highly and then tested in the evolution of the species contained within cave-adapted species. Cave-adapted animals, particularly those that system. Comparisons of divergence times estimated across with close surface relatives, offer the potential for genome-wide taxa within a particular cave, karst system, region, and even

Frontiers of Karst Research 103

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Understanding the Tempo and Mode of Evolution

across regions, will provide insight into whether or not par- this field for several reasons. First, many cave animals show ticular climatic and geologic events have had a strong influ- very strong behaviors that have evolved to maximize survival ence on colonization of the subterranean habitat. With respect in the harsh cave environment. Classic examples are feeding, to the evolution of particular traits, caves also offer the abil- aggression, and reproductive behaviors. Second, behaviors ity to put a timescale on how fast convergence and adaptation have either appeared de novo in the cave environment or have can occur. Understanding when particular species invaded the been changed drastically from a related behavior that still oc- cave habitat places an upper limit on the time of trait acquisi- curs in a surface dwelling ancestor. Many ancestral behaviors tion. Furthermore, by studying species with different degrees also have been lost during adaptation to cave life. Finally, the of cave-adaptation, information can be obtained on which traits existence of closely related surface and cave dwelling species, evolve more rapidly. indeed sometimes divergent populations of the same species, allow powerful genetic approaches to dissect the molecular How Can We Use Cave Forms to Understand basis of evolutionary changes in cave animal behavior via hy- Mutation Rates? bridization. Once again, most of the research in this area has been and will continue to be conducted on Astyanax mexicanus, An emerging theme in evolutionary biology is adaptability po- in which a suite of behavioral changes has been very well de- tential. Understanding the evolution of form requires under- scribed in both the surface and cave forms. However, other standing the generation of the underlying variation on which cave adapted model systems offer the potential to study a wide selection acts. Mutation rates are essential to understanding range of behaviors and perhaps reveal unexpected convergen- both the divergence times of cave species and the evolution ces at the molecular level. of the cave form. Furthermore, although mutation rates are an essential component of many evolutionary theories, they are SUMMARY AND RECOMMENDATIONS difficult to measure. Therefore, few people directly measure mutation rates in their organism of interest, and fewer still in- Recent advances in genetic tools and analyses now allow us to vestigate broader questions such as how mutation rates change take advantage of the model system presented by cave faunas to throughout the course of evolution. Cave species offer interest- address many critical biological issues in unique ways. Cave- ing systems in which it may be possible to address these ques- adapted species present an unparalleled system where interactions. During the course of cave adaptation, generation times tions, mechanisms, and evolutionary forces among opposing increase, affecting the ability of a species to generate genetic traits – constructive and regressed, convergent and divergent, novelty and to respond to environmental changes. However, no selection versus drift – can be investigated. The existence of one has investigated what happens to mutation rates as species numerous, independently colonized karst systems by phyloge- adapt to the subterranean realm. As species become adapted netically diverse species offers a comparative system par excel- to the karst habitat, do mutation rates slow down, speed up, or lence. The main research themes presented here represent a remain unchanged? Species complexes where closely related highly interdisciplinary perspective, including: species have independently invaded multiple cave systems may provide a way to investigate these issues. • Taxonomy: caves species present an ideal system for developing methods for integrating genetic diversity into Can Cave Fauna Help Us Understand the classical taxonomic methods. Evolution of Behavior? • Biodiversity: understanding cryptic diversification and speciation in a system where both seem to be rampant. The evolution of behavior is an emerging discipline ripe for • Comparative genomic studies: genomic investigations in study at the molecular level. What kind and how many genes karst offer the potential to directly measure the proportion govern behavioral change? Are behavioral changes subject to of the genome that is changing relative to invasion of a conventional evolutionary mechanisms? Through what sys- new habitat, convergent versus divergent features, and the tems of the organism do the molecular changes underlying role of selection versus drift in the evolution of the cave behavior act: the nervous system, the endocrine system, the re- form. productive system? Do behavioral changes cause evolutionary • Ecosystem studies: the influence of environmental param- changes or vice versa? Despite widespread current interest in eters such as energy limitation and lack of light as driving these questions, model systems are not routinely available to forces of evolution can be investigated in karst. study the molecular basis or evolution of behavior. Cave adapt- • Behavior: cave adapted animals may provide unique sys- ed animals have the potential to make a major contribution to tem to explore the evolution of behavior at the molecular

104 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Understanding the Tempo and Mode of Evolution

level, a major frontier in biology. For too long, caves have been perceived as highly specialized systems with little to offer the broader fields of biology and evo- For these types of research to be successful, we require: lution. In reality, cave-adapted species offer unrivalled model systems in which broad biological and evolutionary questions • The establishment of at least 2 pairs of surface and cave- can be investigated. The time is overripe to propel cave animals adapted populations or species as model organisms. to the forefront of biological research. • The funds for full genome sequencing and analysis of these model species. • Expanded opportunities for training graduate students in aspects of the ecology of the karst habitat as well as the fields of genome science and molecular evolution.

Frontiers of Karst Research 105

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Focus Group on Karst Resources and Other Applied Issues

Focus Group on Karst Resources and Other Applied Issues

Group Participants: Dorothy J. Vesper and Geary M. Schindel, Leaders. Barry F. Beck, John Van Brahana, Jen Cate, Debra Engler, Ralph Ewers, Joan Falkenberg, Todd Halihan, Peter Idstein, Noel Krothe, Eric Peterson, Laura Toran, George Veni, Elizabeth L. White, and Shannon Williams

MAJOR ISSUES well characterized sites are needed to hone these tools. 5. Changes in land use, particularly urbanization, may affect he major applied issues include the topics of water qual- water quality, water quantity, and collapse features. The ity, water quantity, and process mechanisms related to degree to which this occurs and how it is manifested war- Tgeotechnical problems. The issues of greatest concern rants additional attention. identified by the focus group are discussed below. JUSTIFICATION 1. How and when contaminants are stored and transported in karst systems is not well understood. The complexity of The primary rationale for addressing applied issues in karst is contaminant transport is exacerbated due to the presence to ensure future resource availability, be it water or land, for of trapping mechanisms, sediment transport, and episodic human and ecological populations who reside on karst. The dilution or mobilization. The possibility for rapid injection preservation of these resources for the future requires us to ad- and transport of contaminants, combined with long-term dress unanswered questions related to applied use. storage, results in a setting in which some contaminants may be rapidly flushed through the system whereas others The inconsistent and potentially-rapid throughput, coupled may be sequestered for years or decades, with long-lasting with the physical configuration of the subsurface, result in impacts. Related questions that need to be addressed rela- contaminant transport mechanisms that are not common in tive to contaminant transport are: 1) the movement of all other aquifers. Most past applied research in karst has been phases of materials through the epikarst; 2) the exchange site-specific; therefore, there is a need to address fundamental of water and solutes between the matrix, fracture and con- processes that can be readily transferred to other karst and non- duit permeability zones; and 3) the role of rapidly-chang- karst locations. ing permeability distributions (due to plugging, blowing of plugs, and evolving dissolution pathways). Closely related The potential for rapid impact of karst systems also makes them is the need to understand where contaminants are stored in a good monitoring venue for surface-activity impacts. The the aquifer, particularly if they are immobile. close connection and rapid cycling between the surface and 2. Episodic transport of sediments impacts water quality. The subsurface in karst settings makes them excellent locations to process, and the presence of thresholds for mobility, need study and observe the full range of interdependence of surface/ to be better understood to explain differences between lo- subsurface interactions. The accessibility to the subsurface in cations and for different events in the same location. karst is unique, and our ability to observe the entire system un- 3. The sustainability and specific yield of karst-water re- derground advances the integration of hydrology, biology, and sources continues to be a difficult question to address for chemistry. water-use planning. Greater knowledge of and techniques for determining recharge, storage and long-term yield are Land-use planning, engineering, reliable infrastructure con- needed to better manage water supplies. Tied to this ques- struction, and effective risk management requires a better un- tion are the needs to understand climate-change impacts derstanding of the hazards associated with building on karst. on water availability, the effects of drought, increased wa- Subsidence can be very slow or nearly instantaneous; end ter withdrawal, and the exchange of water into and out of members that requires different strategies. Understanding the storage. underlying processes and controls will help predict risks, aid in 4. Major geotechnical issues relevant to collapse, subsidence alleviating problems, and optimize our karst resources. and infiltration relate to the mechanisms of near-surface erosional processes and the associated soil mechanics. APPROACHES Remote sensing techniques such as geophysical monitoring can improve understanding of these relationships, but Applied issues would be best understood if detailed, integra-

106 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Focus Group on Karst Resources and Other Applied Issues tive (hydrology, chemistry, geophysics, geotechnical, biology) multi-disciplinary studies. Different types of data could be studies were conducted at well-characterized sites. The focus combined and would be available for all researchers. Teams needs to be on fundamental processes rather than site-specific need to incorporate practitioners and regulators as well as re- characterization and delineation. The research sites would searchers so that information gained can be shared to groups facilitate the sharing of techniques for setting up long-term with diverse goals and needs. The inclusion of a field site with monitoring networks, finding the effective tools, and learning multiple contaminants present would remove the access barri- and testing of new equipment. A site users group would offer ers for studying contaminant transport. An infrastructure that advice, workshops, manuals, as well as researchers available provides for long-term monitoring will remove the barriers re- for consultation. lated to collecting temporal-flow variation data and, eventually, include some threshold events in the dataset. A “red team” approach to contamination events would allow researchers to observe rapid responses in karst systems. A Numerous potential sites already exist: the Edwards Aquifer karst red team, similar to those employed by volcano research- research site in San Antonio, Texas, the Savoy Experimental ers, would be ready to mobilize quickly (within several hours) Watershed in Arkansas, Mammoth Cave National Park and ad- when spills occur. Not only would this approach allow team jacent lands, Bowling Green, Kentucky, and the USGS karst members to respond and contain early contamination-plume observatory in Leetown, West Virginia. Research is ongoing at migration, it would also facilitate data collection and research all sites and could be expanded to a larger group of researchers of leading-edge contaminant behavior. Locations that offered via several tasks: publicizing facilities and geologic settings, promise for technology transfer of major processes could be obtaining a small budget for managing logistics, and formal- pursued as long-term research sites. izing access agreements.

MAJOR BARRIERS Use of new tracer techniques and incorporation of more geochemical tools is essential. The most common tracers used Several barriers inhibit ongoing karst research in the applied in karst applied studies are fluorescent dyes that travel in the field: dissolved phase. More development and testing of tracers for sediment transport or different types of contaminants would • Water quantity: Evaluation of storage and water budgets greatly aid our understanding of flow. Tracers that have spe- requires an infrastructure of both wells and springs and in- cific chemical characteristics (solubility, sorption potential) strumentation to collect data on both input and output. The combined with longer-time tracer tests would allow research- record must be long enough to include periods of drought, ers to understand contaminant transport in a more mechanistic major storms, and possible threshold-exceeding impacts. fashion and without the need for contamination. Collabora- • Water quality: There are often legal and access barriers to tion with researchers studying ecosystem function could help; conducting research on contaminated sites. Site specific tracking the “degradation fingerprints” of natural organic mate- information is often not disseminated within the scientific rial may clarify how different types of organic compounds are community due to regulatory or legal concerns by land- cycled and transported. Isotopes are only occasionally used by owners. Given that contamination is an unplanned event, practitioners in applied studies. More collaboration is needed it rarely happens in locations that have long-term records between research geochemists and applied karst hydrogeolo- and infrastructure for research. Availability of analytical gists. instrumentation and support is a barrier for many researchers, particularly for analyzing organic contaminants. Communication between researchers needs facilitation. While • Geotechnical issues: Geotechnical knowledge and tech- final research results are published, knowledge regarding equip- niques are often driven by case-specific needs. Barriers ment and techniques needs to be transferred as well. Establish- include the funding and availability of different types of ment of equipment user groups would help this problem. This karst research sites in different settings that can be utilized could be further solved by testing of equipment at the research for comparison. Being unable to predict when catastrophic field sites. Better communication between field researchers failure will occur limits the study of the collapse process. and modelers would improve both teams understanding of the flow processes and conceptual models. At the research sites the REQUIRED TOOLS field teams could provide calibration datasets for the modelers and the modelers could provide direction on the most important The most important tool to develop for researching applied is- data to collect. sues is a network of research field sites that can be used for

Frontiers of Karst Research 107

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 References for Focus Groups

REFERENCES FOR FOCUS GROUPS

Asmerom, Y. and Polyak, V.J., 2004, Comment on "A test of Dreybrodt, W., 1988, Processes in Karst Systems: Springer- annual resolution in stalagmites using tree rings": Quaternary Verlag, Berlin, 288 p. Research, v. 61, p. 119-121. Drysdale, R., Pierotti, L., Piccini, L., and Baldacci, F., 2004, Asmerom, Y., Polyak, V.J., Burns, S.J., and Rasmussen, J.B.T., Suspended sediments in karst spring waters near Massa (Tus- 2007, Solar forcing of Holocene climate: New insights from cany), Italy: Environmental Geology, v. 40, p. 1037-1050. a speleothem record, southwestern United States: Geology, v. 35(1), p. 1-4. Dussart, L., DuPont, J.P., Zimmerlin, I., Lacroix, M., Saiter, J.M., Junter, G.A., and Jouenne, G., 2003, Occurrence of ses- Banner, J.L., Guilfoyle, A., James, E.W., Stern, L.A., and Mus- sile Pseudomonas oryzihabitans from a karstified chalk aqui- grove, M.L, 2007, Seasonal variations in modern speleothem fer: Water Research, v. 37, p. 1593-1600. calcite growth in central Texas, U.S.A.: Journal of Sedimentary Research, v. 77, p. 615-622. Engel, A.S., Lichtenberg, H., Prange, A., Hormes, J., 2007, Spe- ciation of sulfur from filamentous microbial mats from sulfidic Barrett, M.R., and Williams, W.J., 1989, Occurrence of atrazine cave springs using X-ray absorption near edge spectroscopy: in groundwaters as a result of agricultural use: In Pesticides FEMS Microbiology Letters: v. 269, p. 54-62. in Terrestrial and Aquatic Environments, Blacksburg, VA, Vir- ginia Water Resources Research Center, p. 39-61. Farmer, J.J., and Williams, S.D., 2001, Seasonal and short-term variability in chlorinated solvent concentrations in two karst Beck, B.F., 1986, A generalized genetic framework for the de- springs in middle Tennessee: implications for sampling design: velopment of sinkholes and karst in Florida, U.S.A.: Environ- U.S. Geological Survey Water Resources Investigations Report mental Geology and Water Science, v. 8, p. 5-18. 01-4011, p. 141-149.

Boyer, D.G., and Pasquarell, G.C., 1996, Agricultural land use Field, M.S., 1988, The vulnerability of karst aquifers to chemi- effects on nitrate concentrations in mature karst aquifer: Water cal contamination: EPA report #600/89/008, 13 p. Resources Bulletin, v. 32, p. 565-573. Ford, D.C., and Williams, P.W., 1989, Karst Geomorphology Boyer, D.G., and Pasquarell, G.C., 1999, Agricultural land use and Hydrology,: London, Chapman and Hall, 601 p.. impacts in bacterial water quality in a karst groundwater aquifer: Journal of the American Water Resources Association, v. Frappier, A.B., Sahagian, D., Carpenter, S.J. González, L.A., 35, p. 291-300 and Frappier, B., 2007, A stalagmite proxy record of recent tropical cyclone events: Geology, v. 7 p. 111–114. Carey, M.A., and Lloyd, J.W., 1985, Modeling non-point sources of nitrate pollution of groundwater in the Great Ouse Froehlich, P.N., Klinkhammer, G.P., Bender, M.L., Luedtke, Chalk, U.K.: Journal of Hydrology, v. 78, p. 83-106. N.A., Heath, G.R., Cullen, D., Dauphin, P., Hammond, D., Hartman, B., and Maynard, V., 1979, Early oxidation of organic DeMott, L., Banner, J.L., and Christian, L., 2006, Recent trav- matter in pelagic sediments of the eastern equatorial Atlantic: ertine deposits as records of groundwater processes in urban- suboxic diagenesis: Geochimica et Cosmochimica Acta, v. 43, izing environments: (abstract) Geological Society of America p. 1075-1090. Meeting, Philadelphia, PA. Gabrovsek, F., Romanov, D., and Dreybrodt, W., 2004, Early Doherty, J., 2002, Pest: Model-independent parameter esti- karstification in a duel fracture-aquifer: the role of exchange mation: 4th Ed., Watermark Numerical Computing, variously flow between prominent fractures and a dense net of fissures: paged. Journal of Hydrology, v. 299, p. 45-66.

Drew D., 1996, Agriculturally induced environmental changes Garcia-Fresca, B., Sharp, J.M., Jr., and Pierce, S.A., 2004, in the Burren karst, Western Ireland: Environmental Geology, Hydrogeology in a rapidly growing urban area and urban- v. 28, p. 137–144. enhanced recharge: (abstract) 32nd International Geological Congress, Florence, Italy. Dreybrodt, W., 1981, Kinetics of dissolution of calcite and its application to karstification: Chemical Geology, v. 31, p. 245- Geyer, T., Birk, S. Liedl, R., and Sauter, M., 2008, Quantifica- 269. tion of temporal distribution of recharge in karst systems from spring hydrographs. Journal of Hydrology, v. 348, p. 452-463.

108 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 References for Focus Groups

Gulden, L.E., Rosero, E., Yang, Z.-L., Rodell, M., Jackson, Changes in the isotopic and chemical composition of ground C.S., Niu, G.-Y.,Yeh, P.J.-F., and Famiglietti, J., 2007, Improv- water resulting from a recharge pulse from a sinking stream: ing land-surface model hydrology: Is an explicit aquifer model Journal of Hydrology, v. 211, p. 178-207. better than a deeper soil profile?: Geophysical Research Let- ters, v. 34, L09402, doi 10.1029/2007GL029804 Katz, B., Bohlke, J.K., and Hornsby, H.D., 2001, Timescales for nitrate contamination of spring waters, northern Florida, Harding, K.A., and Ford, D.C., 1993, Impacts of primary defor- USA: Chemical Geology, v. 179, p. 167-186. estation upon limestone slopes in Northern Vancouver-Island, British-Columbia: Environmental Geology, v. 21, p. 137-143. Katz, B.G., 2004, Nitrate contamination in karst ground water, in Encyclopedia of Caves, Culver, D.C. and White, W.B., (eds.), Herczeg, A.L., and Edmunds, W.M., 1999, Inorganic ions as Elsevier Science, Amsterdam, Netherlands, pp. 415-418. tracers, in Cook, P.G., Herczeg, A.L. eds. Environmental Trac- ers in Subsurface Hydrology, Kluwer Academic Press, p. 31- Klimchouk, A.B., Ford, D.C., Palmer, A.N., and Dreybrodt, W., 77. 2000, Speleogenesis: Evolution of Karst Aquifers. National Speleological Society, Huntsville, AL, 527 p. IPCC (Intergovernmental Panel on Climate Change), 1997, IPCC Special Report on The Regional Impacts of Climate Klimchouk, A., 2007, Hypogene Speleogenesis. National Cave Change, An Assessment of Vulnerability (available at: www. and Karst Research Institute Special Paper 1, 106 p. ipcc.ch). Krause, R.E., 1979, Geohydrology of Brooks, Lowndes and IPCC (Intergovernmental Panel on Climate Change), 2001, western Echols Counties, Georgia: U.S. Geological Survey Climate Change 2001: Working Group II: Impacts, Adaptation Water-Resources Investigations Report 78-117. and Vulnerability (available at: www.ipcc.ch). Kuczynskaa, E., Boyera, D.G., and Shelton, D.R., 2003, Com- IPCC (Intergovernmental Panel on Climate Change), 2007, Cli- parison of immunofluorescence assay and immunomagnetic mate Change 2007: The Physical Science Basis, Summary for electrochemiluminescence in detection of Cryptosporidium Policy Makers, Contribution of Working Group I to the Fourth parvum oocysts in karst water samples: Journal of Microbio- Assessment Report of the Intergovernmental Panel on Climate logical Methods, v. 53, p. 17-26. Change (available at: www.ipcc.ch). Labourdette, R., Lascu, I., Mylroie, J. and Roth, M., 2007, Johnson, K. R., Hu, C.Y., Belshaw, N.S. and Henderson, G.M., Process-like modelling of flank margin caves: From genesis to 2006, Seasonal trace-element and stable-isotope variations in burial evolution: Journal of Sedimentary Research, v. 77, p. a Chinese speleothem: The potential for high-resolution pa- 965-979. leomonsoon reconstruction: Earth and Planetary Science Let- ters, v. 244, p. 394-407. Liedl, R., Sauter, M., Hückinghaus, D., Clemens, T., and Teutsch, G., 2003, Simulation of the development of karst aqui- Jones, W.K, Culver, D.C., and Herman, J.S., (eds.), 2003, fers using a coupled continuum pipe flow model. Water Re- Epikarst. Karst Waters Institute Special Publication 9, 160 p. sources Research, v. 39, No. 3, doi 10.1029/2001WR001206. Katz, B.G., Plummer, L.N., and Busenberg, E., Revesz, K.M., Long, A.J., and Putnam, L.D., 2004, Linear model describ- Jones, B.F., and Lee, T.M., 1995, Chemical evolution of ing three components of flow in karst aquifers using 18O data: groundwater near a sinkhole lake, northern Florida: 2. Chemi- Journal of Hydrology, v. 296, p. 254-270. cal patterns, mass-transfer modeling, and rates of chemical reactions: Water Resources Research, v. 31, p. 1565-1584. Long, A.J., and Putnam, L.D., 2006, Translating CFC-based piston ages into probability density functions of ground-water Katz, B.G., Coplen, T.B., Bullen, T.D., and Davis, J.H., 1997, age in karst: Journal of Hydrology, v. 330, p. 735-747. Use of chemical and isotopic tracers and geochemical modeling to characterize the interactions between ground water and Mahler, B.J., 1997, Mobile Sediments in a Karst Aquifer: Aus- surface water in mantled karst: Ground Water, v. 35, p. 1014- tin, TX, University of Texas, Ph.D. Dissertation, 171 p. 1028. Mahler, B.J., Garner, B.D., Musgrove, M., Guilfoyle, A.L., and Katz, B.G., DeHan, R.S., Hirten, J.J., and Catches, J.S., 1997, Rao, M.V., 2006, Recent (2003-05) water quality of Barton Interactions between ground water and surface water in the Su- Springs, Austin, Texas, with emphasis on factors affecting vari- wannee River basin, Florida: Journal of the American Water ability: U.S. Geological Survey Scientific Investigation Report Resources Association, v. 33, p. 1237-1254. 2006-5299, 83 p.. Katz, B.G., Catches, J.S., Bullen, T.D., and Michel, R.L., 1998,

Frontiers of Karst Research 109

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 References for Focus Groups

Mahler, B.J., and Massei, N., 2007, Anthropogenic contami- Mickler, P., Ketcham, R., Colbert, M. and Banner, J.L., 2004b, nants as tracers in an urbanizing karst aquifer: Journal of Con- Application of high-resolution X-ray computer tomography in taminant Hydrology, v. 91, p. 81-106. determining the suitability of speleothems for use in paleocli- matic and paleohydrologic reconstructions: Journal of Cave Mahler, B.J., Personné, J.C., Lods, G.F., and Drogue, C., 2000, and Karst Studies, v.66, p.4-8. Transport of free and particulate-associated bacteria in karst: Journal of Hydrology, v. 238, p. 179-193. Musgrove, M., Banner, J.L., Mack, L.E., Combs, D.M., James, E.W., Cheng, H., and R.L. Edwards, 2001, Geochronology of Mahler, B.J., and Van Metre, P.C., 2000, Occurrence of soluble Late Pleistocene to Holocene speleothems from central Texas: pesticides in Barton Springs, Austin, Texas, in response to a Implications for regional paleoclimate: Geological Society of rain event: U.S. Geological Survey website http://tx.usgs.gov/ America Bulletin, v. 113, p. 1532-1543. reports/dist/dist-2000-02/. Musgrove, M. and Banner, J.L., 2004, Controls on the spatial Martin, J.B. and Dean, R.A., 2001, Exchange of water between and temporal variability of vadose dripwater geochemistry: conduit and matrix in the Floridan aquifer: Chemical geology, Edwards aquifer, central Texas: Geochimica et Cosmochimica v. 179, p. 145-166. Acta, v. 68, p. 1007-1020.

Martin, J.B. and Screaton, F.J., 2001, Exchange of ma- Nicholson, S.E., Tucker, C.J., and Ba, M.B., 1998, Desertifi- trix and conduit water with examples from the Floridan cation, drought, and surface vegetation: An example from the aquifer: In Kuniansky, E.L. (ed.), U.S. Geological Survey West African Sahel: Bulletin of the American Meteorological Water-Resources Investigations Report 01-4011, p. 38-44. Society, v. 79, p. 815-829.

Massei, N., Dupont, J.P., Mahler, B.J., Laignel, B., Fournier, Niu, G.-Y., Yang, Z.-L., Dickinson, R.E., Gulden, L.E., and M., Valdes, D., and Ogier, S., 2006, Investigating transport Su, H., 2007, Development of a simple groundwater mod- properties and turbidity dynamics of a karst aquifer using cor- el for use in climate models and evaluation with GRACE relation, spectral, and wavelet analyses: Journal of Hydrology, data: Journal of Geophysical Research, v. 112, D07103, v. 329, p. 244-257. doi:10.1029/2006JD007522.

Massei, N., Lacroix, M., Wang, H.Q., Mahler, B.J., and Du- O'Reilly, C.E., Bowen, A.B., Perez, N.E., Sarisky, J.P., Shep- Pont, J.P., 2002, Transport of suspended solids from a karstic herd, C.A., Miller, M.D., Hubbard, B.C., Herring, M., Buch- to an alluvial aquifer: the role of the karst/alluvium interface: anan, S.D., Fitzgerald, C.C., Hill, V., Arrowood, M.J., Xiao, Journal of Hydrology, v. 260, p. 88-101. L.X., Hoekstra, M., Mintz, H.D., and Lynch, M.F., 2007, A waterborne outbreak of gastroenteritis with multiple etiologies Massei, N., Wang, H.Q., Dupont, J.P., Rodet, J., and Laignel, among resort island visitors and residents: Ohio, 2004: Clinical B., 2003, Assessment of direct transfer and resuspension of Infectious Diseases, v. 44, p. 506-512. particles during turbid floods at a karstic spring: Journal of Hy- drology, v. 275, p. 109-121. Palmer, A.N., 1991, Origin and morphology of limestone caves: Geological Society of America Bulletin, v. 103, p. 1-21. McMahon, P.B. and Chapelle, F., 1991, Microbial production of organic acids in aquitard sediments and its role in aquifer Palmer, A., 2002, A distinctly European approach to karst hy- geochemistry: Nature, v. 349, p. 233 – 235 drology: Hydrologic Processes, v. 16, p. 2905-2906.

McConnell, J.B. and Hacke, C.M., 1993, Hydrogeology, water Parise, M. and Pascali, V., 2003, Surface and subsurface envi- quality and water resources development potential of the Up- ronmental degradation in the karst of Apulia (southern Italy): per Floridan aquifer in the Valdosta area, south-central Geor- Environmental Geology, v. 44, p. 247-256. gia: U.S. Geological Survey Water Resources Investigations Report, 93-4044. Partin, J., Cobb, K., Adkins, J., Clark, B., and Fernandez, D. (2007). Millennial-scale trends in West Pacific Warm Pool hy- Menon, S., Hansen, J., Nazarenko, L. and Luo, Y., 2002, Cli- drology: Nature, v. 449, p. 452-455. mate effects of black carbon aerosols in China and India: Sci- ence, v. 297, p. 2250-2253. Pasquarell, G.C. and Boyer, D.G., 1996, Herbicides in karst groundwater in Southeast West Virginia: Journal of Environ- Mickler, P.M., Banner, J.L., Stern, L.A., Asmerom, Y., Ed- mental Quality, v. 25, p. 755-765. wards, R.L., and Ito, E., 2004a, Stable isotope variations in modern tropical speleothems: Evaluating equilibrium vs. ki- Parkhurst, D.L. and Appelo, C.A.J., 1999, User's guide to netic isotope effects: Geochimica et Cosmochimica Acta, v. 68, PHREEQC (version 2); a computer program for specia- p. 4381–4393. tion, batch-reaction, one-dimensional transport, and in-

110 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 References for Focus Groups

verse geochemical calculations: U.S. Geological Survey Sauro, U., 1993, Human impact on the karst of the Venetian Water-Resources Investigations 99-4259, 312 p. Fore-Alps, Italy: Environmental Geology, v. 21, p. 115-121.

Plummer L. N., 1977, Defining reactions and mass transfer in Smart, P.L. and Hobbs, S.L., 1986, Characterisation of carbon- part of the Floridan Aquifer: Water Resources Research, v. 13, ate aquifers: A conceptual base: Proceedings of the Environ- p. 801–812. mental Problems in Karst Terranes and Their Solutions Confer- ence, Bowling Green, Kentucky, p. 1-14. Plummer, L.N. and Busenberg, E., 1999, Chlorofluorocarbons, in Cook, P.G., Herczeg, A.L. eds. Environmental Tracers in Treble, P.C., Chappell, J., Gagan, M.K., McKeegan, K.D., and Subsurface Hydrology, Kluwer Academic Press, p. 441-478. Harrison, T.M., 2005, In situ measurement of seasonal delta O-18 variations and analysis of isotopic trends in a modern spe- Plummer, L.N., Prestemon, E.C., and Parkhurst, D.L., 1992, leothem from southwest Australia: Earth and Planetary Science NETPATH; an interactive code for interpreting NET geo- Letters, v. 233, p. 17-32. chemical reactions from chemical and isotopic data along a flow PATH: Proceedings of the 7th international symposium Vaute, L., Drogue, C., Garrelly, L., and Ghelfenstein, M., 1997, on water-rock interaction, Volume 1, Low temperature environ- Relations between the structure of storage and the transport of ments, p. 239-242 chemical compounds in karstic aquifers: Journal of Hydrology, v. 205, p. 221-238. Plummer, L.N., Busenberg, E., McConnell, J.B., Drenkard, S., Schlosser, P., and Michel, R., 1998, Flow of river water into Vesper, D.J., Loop, C.M., and White, W.B., 2001, Contaminant a karstic limestone aquifer. 1. Tracing the young fraction in transport in karst aquifers: Theoretical and Applied Karstology, groundwater mixtures in the Upper Floridan aquifer near Val- v. 13-14, p. 101-111. dosta, Georgia: Applied Geochemistry, v. 13, pp. 995-1015. Wang, S.J., Liu, Q.M., and Zhang, D.F., 2004, Karst rocky de- Polyak, V.J., Rasmussen, J.T., and Asmerom, Y., 2004, Pro- sertification in southwestern China: Geomorphology, landuse, longed wet period in the southwestern United States through impact, and rehabilitation: Land Degradation and Develop- the Younger Dryas: Geology, v. 32, p. 5-8. ment, v. 15, p. 115-121.

Quinlan, J.F., 1989, Ground-water monitoring in karst terranes: Wang, X., Auler, A.S., Edwards, R.L., Cheng, H., Ito, E., and Recommended protocols and implicit assumptions: U.S. Envi- Solheid, M., (2006). Interhemispheric anti-phasing of rainfall ronmental Protection Agency EPA 600/X-89/050. during the last glacial period. Quaternary Science Reviews, v. 25, p. 3391-3403. Rasmussen, J.B.T., Polyak, V.J. and Asmerom, Y., 2006, Evi- dence for Pacific-modulated precipitation variability during the Wang, Y.J., Cheng, H., Edwards, R.L., An, Z.S., Wu, J.Y., Shen, late Holocene from the southwestern USA: Geophysical Re- C.C., and Dorale, J.A., 2001, A high-resolution absolute-dated search Letters, v. 33: L08701, doi:10.1029/2006GL025714. Late Pleistocene monsoon record from Hulu Cave, China: Sci- ence, v. 294, p. 2345-2348. Romanov, D., Gabrovsek, F., and Dreybrodt, W., 2003, The impact of hydrochemical boundary conditions on the evolution Wicks, C., Kelley, C., and Peterson, E., 2004, Estrogen in a of limestone karst aquifers: Journal of Hydrology, v. 276, p. karstic aquifer: Ground Water, v. 42, p. 384-389. 240-253. Williams, P.W., 1993, Environmental change and human im- Rostad, C.E., Leenheer, J.A., Katz, B.G., Martin, B.S., and pact on karst terrains: An introduction, in Williams, P. W. (ed.), Noyes, T.I., 2000, Characterization and disinfection by-product Karst Terrains: Environmental changes and human impact, formation potential of natural organic matter in surface and Catena Suppl. v. 25, p. 1-20. ground waters from northern Florida: in Barrett, S.E., Krasner, S.W., and Amy, G.L., (eds.), Natural Organic Matter and Dis- Williams, P.W., King, D.N.T., Zhao, J.X., and Collerson, K.D., infection Byproducts, American Chemical Society Symposium 2005, Late pleistocene to Holocene composite speleothem Series 761, Ch.11, p.154-172. O-18 and C-13 chronologies from South Island, New Zealand- did a global Younger Dryas really exist?: Earth and Planetary Ryan, M., and Meiman, J., 1996, An examination of short- Science Letters, v. 230, p. 301-317. term variations in water quality at a karst spring in Kentucky: Ground Water, v. 34, p. 23-30. Williams, P.W. (ed.), 1993, Karst Terrains: Environmental Change and Human Impact: Catena Supplement 25, Catena Sasowsky, I.D., Feazel, C.T., Mylroie, J.E., Palmer, A.N. and Verlag, Cremlingen-Destedt, Germany, 268 p. Palmer, M.V., (eds.) 2008, Karst from Recent to Reservoirs. Karst Waters Institute Special Publication 14, 221 p. Wolfe, W.J. and Williams, S.D., 1999, Soil gas screening for

Frontiers of Karst Research 111

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 References for Focus Groups

chlorinated solvents at three contaminated karst sites in Ten- Zuber, A., Weise, S.M., Motyka, J., Osenbrück, K., and nessee: Ground Water Monitoring and Remediation, v. 22, p. Różański, K., 2004, Age and flow pattern of groundwater in 91-99. a Jurassic limestone aquifer and related Tertiary sands derived from combined isotope, noble gas and chemical data: Journal Yin, G.G., Kookana, R.S., and Ru, Y.J., 2002, Occurrence and of Hydrology, v. 286, p. 87-112. fate of hormone steroids in the environment: Environment In- ternational, v. 28, p. 545-551.

112 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Appendix: Workshop Participants

Appendix: Workshop Participants

Focus Group on Karst Resources and Applied Issues

Barry Beck PE LaMoreaux and Associates Noel Krothe 106 Administration Rd., Ste. 4 Indiana University Oak Ridge, TN 37830 1211 South Walnut St. [email protected] Bloomington, IN 47404 [email protected] John V. Brahana University of Arkansas Eric Peterson Department of Geosciences Illinois State University 113 Ozark Hall Department of Geography-Geology Fayetteville, AR 72701 Campus Box 4400 [email protected] Normal, IL 61790 [email protected] Jen Cate Texas A&M University Geary Schindel 3102 Cove View Blvd. #P203 Edwards Aquifer Authority Galveston, TX 77554 11310 Whisper Dawn [email protected] San Antonio, TX 78230 [email protected] Debra Engler SAWS Laura Toran 2800 US Hwy. 281 N Temple University San Antonio, TX 78205 Department of Geology [email protected] 1901 N. 13th St. Philadelphia, PA 19122 Ralph Ewers [email protected] Ewers Inc. EKV 160 Redwood Dr. George Veni Richmond, KY 40475 National Cave and Karst Research Institute [email protected] 1400 Commerce Dr. Carlsbad, New Mexico 88220 Joan Falkenberg [email protected] SAWS 2800 US Hwy. 281 N Dorothy Vesper San Antonio, TX 78205 West Virginia University [email protected] Department of Geology and Geography Morgantown, WV 26506-6300 Todd Halihan [email protected] OSU Stillwater School of Geology Elizabeth L. White 105 NRC Hydrologic Investigations Stillwater, OK 74078 4538 Miller Rd. [email protected] Petersburg, PA 16669-2711 [email protected] Peter Idstein EWC Shannon Williams 971 Villa Dr. Apt. 27 USGS-Tennessee Richmond, KY 40475 640 Grassmere Park, Ste. 100 [email protected] Nashville, TN 37211 [email protected]

Frontiers of Karst Research 113

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Appendix: Workshop Participants

Focus Group on Hydrology and Hydrogeology

Matt Covington William K. Jones UCSC Karst Waters Institute 74 Barnes Ct. 408 PO Box 356 Standford, CA 94305 Warm Springs, VA 24484 [email protected] [email protected]

Lee Florea Todd Kincaid US Geological Survey Hazlett-Kincaid Inc. 3110 SW 9th Ave. 27 Keystone Ave. Ft. Lauderdale, Fl 33315 Reno, NV 89503 [email protected] [email protected]

Franci Gabrovsek PJ Moore Karst Research Inst., Slovenia University of Florida Titov Trg 2 PO Box 112120 6230 Postojna, Slovenia Gainesville, FL 32611 [email protected] [email protected]

Yongli Gao John Mylroie East Tennessee State University Mississippi State University Box 70652 PO Box 5448 East TN State Univ., TN 37614 Department of Geociences [email protected] Mississippi State, MS 39762 [email protected] Ronald Green Southwest Research Inst. Ira Sasowsky 6220 Culebra The University of Akron San Antonio, TX 78232 Department of Geology and Environmental Science [email protected] Akron, OH 44325-4101 [email protected] Jason Gulley University of Florida Martin Sauter PO Box 112120 University of Göttingen Gainesville, Fl 32611 Goldschmidtstr. 3, 37077 [email protected] Göttingen Germany [email protected] Russell Harmon US Army Research Office Elizabeth Screaton PO Box 12211 Geological Sciences Research Triangle Park, NC 27709-2211 University of Florida [email protected] Gainesville, FL 32611 [email protected] Ellen Herman Bucknell University Carol Wicks 224 O'Leary Center University of Missouri, Columbia Department of Geology 101 Geology Bldg. Lewisburg, PA 17837 Columbia, MO 65211 [email protected] [email protected]

Pierre-Yves Jeannin Director of the Swiss Institute of Karst PO Box 818 CH-2301 La Choux-de-fonds, Switzerland [email protected]

114 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Appendix: Workshop Participants

Focus Group on Geochemistry and Climate

Jay Banner Jon Martin University of Texas - Austin University of Florida 3604 Crowncrest Dr. PO Box 112120 Austin TX 78759 Gainesville, FL 32611 [email protected] [email protected]

Liza Colucci MaryLynn Musgrove City of Austin USGS 1824 S 1H 35 Apt. 360 8027 Exchange Dr. Austin, TX 78754 Austin, TX 78754 [email protected] [email protected]

Brian Cowan Jud Partin University of Texas - Austin Georgia Tech 10926 Jollyville Rd. #422 311 Ferst Dr. Austin, TX 78754 Atlanta, GA 30332 [email protected] [email protected]

Amy Frappier Jessica Rasmussen Boston College University of Texas - Austin 140 Commonwealth Ave. 8700 Brodie Ln. #922 Devlin Hall 213 Austin, TX 78745 Chestnut Hill, MA 02467 [email protected] [email protected] Corinne Wong Cara Gentry University of Texas - Austin University of Florida 1824 S 1H 35 Apt. 360 PO Box 112120 Austin, TX 78754 Gainesville, FL 32611 [email protected] [email protected] William B. White Brian Katz Penn State USGS Mat. Res. Lab. Bldg. 2010 Levy Ave. University Park, PA 16802 Tallahassee, FL 32310 [email protected] [email protected]

Andrew Long USGS Water Science Center 1608 Mt. View Rd. Rapid City, SD 57702 [email protected]

Frontiers of Karst Research 115

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Appendix: Workshop Participants

Focus Group on Geomicrobiology

Annette Engel Dan Jones Louisiana State University Penn State E235 Hove-Russell Geosciene Complex 242 Deike Bldg. Department of Geology and Geophysics Geosciences Department Baton Rouge, LA 70903 University Park, PA 16802 [email protected] [email protected]

Marcus Gary Jenn Macalady USGS Penn State University 8027 Exchange Dr. Geosciences Department Austin, TX 78754 University Park, PA 16802 [email protected] Brett Gonzalez Texas A&M University Diana Northup 5007 Ave. U University of New Mexico, Biology Galveston, TX 77551 MSCO3 2020 [email protected] Albuquerque, NM 87131 [email protected] Juan Gonzalez Instituto Resources Naturales Agrobiol (CSIC) John Spear Avda Reina Mercedes 10 Colorado School of Mines 41012 - Sevilla, Spain 1500 Illinois St. [email protected] Golden, CO 80401 [email protected] Elena Hutchens UCD Ireland Mike Spilde School of Biological and Environmental Science University of New Mexico Belfield, UCD, Ireland Inst. of Meteorites [email protected] MSC 03-2050 1 University of New Mexico Albuquerque, NM 87131 [email protected]

Focus Group on Ecosystem Science

Daniel Fong Robert Payn American University Colorado School of Mines Department of Biology GEGN Department Washington, DC 20016 1516 Illinois St. [email protected] Golden, CO 80401 [email protected] Lara Hinderstein Texas A&M University Kevin Simon Department of Marine Biology University of Maine 5007 Ave. U School of Biology and Ecology Galveston, TX 77551 Orono, ME 04469 [email protected] [email protected]

Bridget Maloney Michael Venarsky Texas A&M University University of Alabama Department of Marine Biology Dept. of Biological Sciences 5007 Ave. U [email protected] Galveston, TX 77551 [email protected]

116 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Appendix: Workshop Participants

Frank Wilhelm Department of Zoology Southern Illinois University 1125 Lincoln Dr. Carbondale, IL 62901-6501 [email protected]

Focus Group on Subterranean Biodiversity

Penny Boston John Holsinger NMT and NCKRI Department of Biological Sciences EES Department Old Dominion University 801 Leroy Pl. Norfolk, VA 23529 Socorro, NM 87801 [email protected] [email protected] Tom Iliffe Mary Christman Texas A&M University University of Florida 5007 Ave. U PO Box 110339 Galveston, TX 77551 Gainesville, FL 32611 [email protected] [email protected] Jean Krejca Valerie Collins Zara Environmental Texas DOT 118 W Goforth Rd. 4610 NW Loop 410 Buda, TX 78610 San Antonio, TX 78229 [email protected] [email protected] Jerry Lewis David Culver Lewis and Associates Department of Biology 17903 State Rd. 60 American University Borden, IN 47106 4400 Massachusetts Ave. [email protected] Washington DC 20016 [email protected] Kathleen O'Conner Travis Co. Natural Resources Jim Godwin 7007 Chinook Dr. Alabama Nat. Heritage Prog. Austin, TX 78736 AUEI [email protected] Auburn University, AL 36849 [email protected] Tanja Pipan Karst Research Inst., Slovenia Horton Hobbs ZRC Sazu, Titov TRG 2 Wittenberg University 6230 Postojna, Slovenia Department of Biology [email protected] PO Box 720 Springfield, OH 45501 Katie Schneider [email protected] University of Maryland 1204 Biol-Psych Bldg. Brian Holmes College Park, MD 20742 TxDOT [email protected] 112 E. Riverside Austin, TX Steve Taylor [email protected] Illinois Natural History Survey 18165 Oak St. Champaign, IL 61821 [email protected]

Frontiers of Karst Research 117

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13 Appendix: Workshop Participants

Maja Zagmajster Department of Biology University of Ljubljana Viena Pot 111, SL-1000 Ljubljana, Slovenia [email protected]

Focus Group on Biological Evolution

Meridith Protas UC Berkeley Megan Porter 1249 Spruce St. University of Maryland, Baltimore County Berkeley, CA 94709 Department of Biological Sciences [email protected] 1000 Hilltop Circle Baltimore, MD 21250 Tristan Lefebure [email protected] Population Medicine and Diagnostic Sciences Cornell University William Jeffery Ithaca, NY 14850 University of Maryland, College Park [email protected] Department of Biology College Park, MD 20742 Katharina Dittmar De La Cruz [email protected] SUNY Buffalo 109 Cooke Hall Ben Hutchins Dept. of Biological Sciences American University Washington DC Buffalo, NY 14260 4105 Wisconsin Ave. NW, Apt. 111 [email protected] Washington, DC 20016 [email protected] Pierre Paquin 12, Chemin Saxby Sud, Shefford, Quebec Canada, J2M 1S2 [email protected]

118 Frontiers of Karst Research

Ka r s t Wa t e r s In s t i t u t e Sp e c i a l Pu b l i c a t i o n 13